motorist behaviour at railway level crossings: the … · crossings than urban older drivers, with...
TRANSCRIPT
MOTORIST BEHAVIOUR AT RAILWAY LEVEL CROSSINGS: THE
PRESENT CONTEXT IN AUSTRALIA
Angela Wallace
Bachelor of Nursing (Australian Catholic University)
Graduate Diploma in Education (Griffith University)
Master of Public Health (Queensland University of Technology)
A thesis submitted for the Degree of Doctor of Philosophy,
Queensland University of Technology,
Centre for Accident Research and Road Safety – Queensland,
Brisbane, Australia.
2008
i
KEY WORDS
Level crossings, railway collisions, road users, heavy vehicles, older drivers,
younger drivers, human factors, risk management, road safety, educational interventions.
ii
ABSTRACT
Railway level crossing collisions in Australia are a major cause of concern for
both rail and road authorities. Despite the fact that the number of railway crash fatalities
in Australia has fallen in recent years, level crossing collisions constitute a significant
proportion of the national rail toll. Although rail transport is presently one of the safest
forms of land transport, collisions at level crossings are three times more likely to
involve fatalities as compared to all other types of road crashes (Afxentis, 1994). With
many level crossing fatalities and injuries resulting in coronial inquests, litigation and
negative media publicity, the actions of rail and road infrastructure providers and the
behaviour of motorists, pedestrians and rail users, come under close scrutiny.
Historically, research in this area has been plagued by the rail/road interface and the
separation of responsibilities between rail and road authorities reflecting the social and
political context in which they are contained. With the recent rail reform in Australia,
safety at level crossings has become a key priority area.
Accordingly, there is a need to better understand the scope and nature of motorist
behaviour at level crossings, in order to develop and implement more effective
countermeasures for unsafe driving behaviour. However, a number of obstacles have
hindered research into the area of level crossing safety. As with many road crashes, the
contributing causes and factors are often difficult to determine, however a recent
investigation of fatal collisions at level crossings supports the notion that human fault is
a major contributor (Australian Transport Safety Bureau, 2002a). Additionally, there is
a lack of reliable data available relating to the behavioural characteristics and
perceptions of drivers at level crossings. Studies that do exist have lacked a strong
theoretical base to guide the interpretation of results.
Due to the lack of financial viability of continuing to approach risk management
from an engineering perspective, the merits of human factor research need to be
examined for suitability. In Australia, there has been considerable recognition regarding
the importance of human factor approaches to level crossing safety (Australian
Transport Council, 2003). However, little attempt has been made by authorities to
scientifically develop and measure the effectiveness of road safety educational
iii
interventions. Therefore, there exists a significant need for developing targeted road
safety educational interventions to improve current risk management solutions at level
crossings.
This research program is the first of its kind in investigating motorist behaviour at
level crossings and the measuring the effectiveness of educational interventions for
improving driving safety. Although other ‘educational’ campaigns exist in this field, no
campaign or intervention has been guided by empirical research or theory. This thesis
adopted a multidisciplinary approach to theory, reviewing perspectives from
psychology, sociology and public health to explain driver behaviour at level crossings.
This array of perspectives is necessary due to the variety of behaviours involved in
collisions and near-misses at level crossings. The motivation underlying motorist
behaviour determines to a large extent how successful behaviour change strategies (e.g.
educational interventions) may be. Fishbein’s Integrated Model of Behaviour Change
(IM) based largely on the health belief model, theory of reasoned action and theory of
planned behaviour (Fishbein, 2000), assisted in the planning and development of a ‘one-
off’ targeted educational intervention specific for three different road user groups and in
questionnaire development to ascertain the present context of motorist behaviour at level
crossings. As no known research has been conducted that utilizes any psychosocial
model to explain or predict level crossing behavior within different road user groups,
this research program used this model as an exploratory tool rather than a tool to asses
the model’s capacity in explaining such behaviour. The difference between this model
and others is the inclusion of two important constructs in driving: skills (or abilities) and
environmental factors. Fishbein (2003) suggests that the model recognises the lack of
skills (or abilities) and/or environmental constraints may prevent a person from acting
on their intentions, in light of the fact that intention is viewed as the primary determinant
of behaviour. While the majority of behaviour change theories are limited by a range of
conceptual and contextual factors (Parker, 2004), the IM was used to assist this research
program as it appeared to be the most applicable model to examining level crossing
safety.
A variety of data collection methods were used in this research program as much
of what is currently known about level crossing collisions is derived from coroner’s
iv
findings and statistics. The first study (Study One) was designed to extend this
knowledge by undertaking a more thorough examination of contributing factors to level
crossing crashes and the road user groups at risk. This study used the method of
‘triangulation’ (i.e. combining research methods to give a range of perspectives)
whereby both qualitative (focus groups) and quantitative (modified Delphi technique)
research designs were utilised (Barbour, 1999, Bryman, 1992). With the discipline of
road safety research requiring methodological strategies that will enhance efforts to
conceptualise the multi-faceted nature of motorist behaviour at level crossings, this
application provided the robustness required. Results from the Delphi technique
indicated that older, younger and heavy vehicle drivers are considered to be three of the
highest risk road user groups by experts in the field. For the older driver group, experts
agreed that errors in judgment were the most important issue for this group when driving
at level crossings. Risk taking by younger drivers, such as trying to beat the train across
the crossing, was viewed as the central issue for the younger driver group. Like the
younger driver group, a concern by experts with the heavy vehicle group was intentional
risk taking at level crossings. However, experts also rated the length of heavy vehicles a
major concern due to the possibility of a truck over-hanging a crossing.
Results from focus groups with train drivers in Study One indicated that there are
unique problems associated with crossings in rural/regional areas compared to urban
areas. The metropolitan train drivers generally experienced motorist behaviour at active
crossings with flashing lights and boom gates while the regional train drivers
experienced behaviours at active crossings with boom gates, crossings with lights only
and passive crossings with stationary signs. In the metropolitan train driver group,
experiences of motorist behaviour at level crossings included: motorists driving around
boom gates, getting stuck under boom gates, queuing over congested crossings and
driving through the crossing after the red lights commence flashing. The behaviour of
motorists driving around boom gates was noted to occur quite regularly. The majority
of metropolitan train drivers reported that it was a common occurrence for motorists to
drive through a crossing when the lights are flashing both before and after the booms
were activated and some crossings were named as ‘black spots’ (locations where
motorists repeatedly violate the road rules). Vehicles protruding into the path of the
v
train and motorists entering congested crossings and then panicking and driving
backwards into the boom gates were also mentioned. Regional train drivers indicated
that motorists not stopping or giving way to trains is a continual problem at passively
controlled crossings (i.e. no boom gates or flashing lights). Regional train drivers
generally agreed that the majority of motorists obey protection systems; however some
motorists drive through flashing lights or drive around boom gates. Other high risk
behaviours included motorists attempting to beat the train across the crossing, speeding
up to go through flashing lights, and general risk taking by younger drivers in particular.
Motorists not allowing enough time to cross in front of the train or hesitating (stop-
starting) at crossings were also noted to be at high risk. There was a general perception
by regional train drivers that motorists are unable to judge the speed and distance of an
approaching train to determine a safe gap during which to cross. Local motorists were
also reported to be a problem at level crossings for regional train drivers. A theme
common to regional and metropolitan train drivers was the risk of catastrophic consequence
associated with level crossing collisions. The reasons given for this were the threat of
derailment, serious property damage, the high risk of a fatality, personal injury and, most
earnestly, the potential for enduring psychological consequences. Drivers uniformly spoke
about the continual fear they had of being involved in a collision with a heavy vehicle, and
many spoke of the effects that such collisions had on train drivers involved. For this reason,
train drivers were said to consider any near-miss incident involving trucks particularly
serious.
The second study undertaken as part of this research program (Study Two),
involved formative research as part of the planning, development and delivery of
behavioural interventions for each of the three road user groups identified in Study One.
This study also used both qualitative and quantitative data collection methods to provide
methodological triangulation and ensure reliability of the data. The overall objective of
the qualitative data collection was to obtain rich data using a qualitative mode of inquiry,
based on the key variables of attitudes, norms, self-efficacy (perceived behavioural
control), perceived risk, environmental constraints and the skills/abilities of drivers. The
overall objective of the quantitative data collection was to prioritise the issues identified
in order to direct and allocate project resources for intervention planning, development
and delivery. This combined recruitment strategy was adopted as it was an appropriate
vi
and practical data collection strategy within the qualitative and exploration
methodology. Information obtained from each of the groups was critical in assisting,
guiding, and identifying priority areas for message and material development. The use
of focus groups and one-on-one interviews provided insights into why drivers think or
do what they do at level crossings. The qualitative component of this study found that
for the older driver group, regional drivers hold a greater perception of risk at level
crossings than urban older drivers, with many recalling near-misses. Participants from
the urban older driver group indicated that level crossings are not as dangerous as other
aspects of driving, with many participants being doubtful that motorists are killed while
driving at level crossings. Both urban and regional younger drivers tended to hold a low
perception of risk for driving at level crossings, however many participants reported
having great difficulty in judging the distance a train is from a crossing. Impatience for
waiting at level crossings was reported to be the major reason for any risk taking at level
crossings in the younger driver group. Complacency and distraction were viewed by
heavy vehicle participants as two of the major driver factors that put them at risk at level
crossings, while short-stacking (when the trailer of the truck extends onto the crossing),
angle of approach (acute or obtuse angle) and lack of advance warning systems were
seen as the major engineering problems for driving a truck at level crossings. The
quantitative component of this study involving research with train drivers found that at
the aggregate train driver level, it is apparent that train drivers consider motorists’
deliberate violations of the road rules and negligently lax approach to hazard detection
as the predominant causes of dangerous driving at level crossings. Experts were
observed to rank risk taking behaviours slightly lower than train drivers, although they
agreed with train drivers that ‘trying to beat the train’ is the single most critical risk
taking behaviour observed by motorists.
The third study (Study Three) involved three parts. The aim of Part One of this
study was to develop targeted interventions specific to each of the three road user groups
by using Fishbein’s theoretical model (Integrated Model of Behaviour Change) as a
guide. The development of interventions was originally seen as being outside of the
scope of this project, however it became intertwined in questionnaire development and
thus deemed to be within the realms of the current mode of inquiry. The interventions
vii
were designed in the format of a pilot radio road safety advertisement, as this medium
was found to be one of the most acceptable to each of the road user groups as identified
in the formative research undertaken in Study Two. The interventions were used as a
‘one-off’ awareness raising intervention for each road user group. Part Two involved
the investigation of the present context of unsafe driving behaviour at level crossings.
This second part involved the examination of the present context of motorist behaviour
at level crossings using key constructs from Fishbein’s Integrated Model of Behaviour
Change (IM). Part Three involved trialing a pilot road safety radio advertisement using
an intervention and control methodology. This part investigated the changes in pre and
post-test constructs including intentions, self-reported behaviour, attitudes, norms, self-
efficacy/perceived behaviour control, perceived risks, environment constraints and
skills/ability. Results from this third study indicated that younger drivers recognise that
level crossings are potentially a highly dangerous intersection yet are still likely to
engage in risk taking behaviours. Additionally, their low levels of self-efficacy in
driving at level crossings pose challenges for developing interventions with this age
group. For the older driver sample, this research confirms the high prevalence of
functional impairments such as increasing trouble adjusting to glare and night-time
driving, restricted range of motion to their neck and substantial declines in their hearing.
While factors contributing to the over-representation of older drivers in collisions at
level crossings are likely to be complex and multi-faceted, such functional impairments
are expected to play a critical role. The majority of heavy vehicle drivers reported
driving safely and intending to drive safely in the future, however, there is a sub-set of
drivers that indicate they have in the past and will in the future take risks when
traversing crossings. Although this sub-set is relatively small, if generalised to the larger
trucking industry it could be problematic for the rail sector and greater public alike.
Familiarity was a common factor that was found to play a role in driving intention at
level crossings for all three road user groups. This finding supports previous research
conducted by Wigglesworth during the 1970’s in Australia (Wigglesworth, 1979).
Taken together, the results of the three studies in this research program have a
number of implications for level crossing safety in Australia. Although the ultimate goal
to improve level crossing safety for all motorists would be to have a combination of
viii
engineering, education and enforcement countermeasures, the small number of fatalities
in comparison to the national road toll limits this. It must be noted though that the
likelihood of creating behavioural change would be increased if risk taking at level
crossings by all motorists was detected and penalised, or alternatively, if perceptions of
such detection were increased. The instilling of fear in drivers with the threat of
punishment via some form of sanction can only be achieved through a combination of a
mass media campaign and increasing police presence. Ideally, the aim would be to
combine fear of punishment with the guilt associated with the social non-acceptability of
disobeying road rules at level crossings. Such findings have direct implications for
improving the present context of motorist behaviour at level crossings throughout
Australia.
ix
TABLE OF CONTENTS
KEY WORDS....................................................................................................................i
ABSTRACT..................................................................................................................... ii
TABLE OF CONTENTS................................................................................................ix
LIST OF FIGURES .......................................................................................................xv
LIST OF TABLES ........................................................................................................xvi
LIST OF APPENDICES ..............................................................................................xix
GLOSSARY OF ACRONYMS AND TERMS............................................................xx
STATEMENT OF ORIGINAL AUTHORSHIP ..................................................... xxii
ACKNOWLEDGMENTS ......................................................................................... xxiii
CHAPTER ONE: INTRODUCTION TO THE THESIS ...........................................1
1.1 INTRODUCTION ..........................................................................................................2
1.2 THE RESEARCH AREA .................................................................................................3
1.3 RATIONALE FOR THE RESEARCH.................................................................................4
1.4 THEORETICAL FRAMEWORK .......................................................................................5
1.5 RESEARCH QUESTIONS AND OBJECTIVES ....................................................................6
1.6 SCOPE OF RESEARCH ..................................................................................................8
1.7 THESIS OUTLINE .........................................................................................................8
1.8 SUMMARY..................................................................................................................9
CHAPTER TWO: LITERATURE REVIEW............................................................11
2.1 INTRODUCTION ........................................................................................................12
2.2 AUSTRALIA’S RAIL INDUSTRY ..................................................................................12
2.2.1 Economic importance...................................................................................12
2.2.2 Rail activities................................................................................................13
2.2.3 Reform agenda .............................................................................................13
2.2.4 Rail safety management ...............................................................................14
2.2.5 National strategy initiative ...........................................................................16
x
2.2.6 Australasian railway level crossing behavioural strategy ............................17
2.2.7 Roles and responsibilities for level crossing safety .....................................18
2.3 LEVEL CROSSING PROTECTION SYSTEMS..................................................................19
2.3.1 Types of protection systems.........................................................................19
2.3.2 Definitions....................................................................................................21
2.3.3 Australian standards .....................................................................................22
2.3.4 International comparison of protection systems ..........................................24
2.4 LEVEL CROSSING STATISTICS IN AUSTRALIA ............................................................25
2.4.1 Data recording..............................................................................................25
2.4.2 Incidence of occurrences..............................................................................26
2.4.3 Incidence of fatalities ...................................................................................27
2.4.4 Cost of collisions..........................................................................................30
2.4.5 International statistical comparison..............................................................34
2.5 ENGINEERING MEASURES .........................................................................................36
2.5.1 Overview of engineering measures..............................................................36
2.5.2 Signage and markings ..................................................................................38
2.5.3 Rumble strips ...............................................................................................38
2.5.4 Sight distances..............................................................................................40
2.5.5 Information handling zones .........................................................................42
2.5.6 Crossing surface ...........................................................................................43
2.5.7 Approach speed limits..................................................................................43
2.5.8 Boom barriers...............................................................................................44
2.5.9 Conspicuity of trains ....................................................................................44
2.5.10 Train horns ...................................................................................................50
2.5.11 Low-cost warning systems...........................................................................51
2.5.12 Vehicle dimensions and performance ..........................................................53
2.5.13 Crossing closure ...........................................................................................54
2.5.14 Risk based scoring systems..........................................................................55
2.6 LAWS AND ENFORCEMENT .......................................................................................56
2.6.1 Overview of enforcement measures.............................................................56
2.6.2 Level crossing laws in Australia ..................................................................57
xi
2.6.3 Enforcing laws at level crossings.................................................................57
2.6.4 Fines and penalties .......................................................................................59
2.7 EDUCATION, PUBLICITY AND TRAINING....................................................................60
2.7.1 Overview......................................................................................................60
2.7.2 Driver training..............................................................................................61
2.7.3 Publicity .......................................................................................................61
2.7.4 Education programs and interventions.........................................................64
2.7.5 Level crossing safety programs and interventions .......................................66
2.7.5 Limitations of current level crossing safety programs.................................71
2.8 HUMAN FACTORS CONTRIBUTING TO COLLISIONS ....................................................72
2.8.1 Overview of accident causation ...................................................................72
2.8.2 Driver interaction with different protection systems ...................................77
2.8.3 Perception of risk .........................................................................................79
2.8.4 High risk behaviours and risk taking ...........................................................80
2.8.5 Familiarity ....................................................................................................86
2.8.6 Approach behaviour .....................................................................................87
2.8.7 Poor knowledge of road rules ......................................................................88
2.8.8 Distraction....................................................................................................89
2.8.9 Attentional blindness....................................................................................91
2.8.10 Hypovigilance and fatigue ...........................................................................91
2.8.11 Speeding.......................................................................................................92
2.8.12 Alcohol and drugs ........................................................................................95
2.8.13 Gender ..........................................................................................................96
2.9 RESEARCH QUESTIONS AND OBJECTIVES ..................................................................97
2.10 SUMMARY................................................................................................................98
CHAPTER THREE: THE IMPORTANCE OF THEORY FOR IMPROVING
MOTORIST BEHAVIOUR AT LEVEL CROSSINGS...........................................100
3.1 INTRODUCTION ......................................................................................................101
3.2 THE ROLE OF THEORY IN ROAD SAFETY STUDIES ....................................................101
3.3 THE INTENTION-BEHAVIOUR RELATION..................................................................102
3.4 INTEGRATED MODEL OF BEHAVIOUR CHANGE........................................................103
xii
3.4.1 Development of the model .........................................................................103
3.4.2 Criticisms of the model ..............................................................................105
3.4.3 Application to level crossing behaviour.....................................................106
3.5 SUMMARY..............................................................................................................106
CHAPTER FOUR: HIGH RISK AND VULNERABLE ROAD USERS AT
LEVEL CROSSINGS ..................................................................................................107
4.1 INTRODUCTION ......................................................................................................108
4.2 REVIEW OF EXISTING DATA ....................................................................................108
4.3 STUDY AIMS AND RESEARCH QUESTIONS................................................................109
4.4 METHOD.................................................................................................................110
4.4.1 Overview....................................................................................................110
4.4.2 Delphi technique ........................................................................................110
4.4.3 Focus groups ..............................................................................................115
4.4.4 Ethical clearance ........................................................................................119
4.5 RESULTS.................................................................................................................119
4.5.1 Delphi technique ........................................................................................119
4.5.2 Focus groups ..............................................................................................121
4.6 DISCUSSION............................................................................................................124
4.6.1 Study limitations ........................................................................................124
4.7 SUMMARY..............................................................................................................125
CHAPTER FIVE: PLANNING AND DEVELOPMENT OF INTERVENTIONS
FOR EACH ROAD USER GROUP...........................................................................127
5.1 INTRODUCTION ......................................................................................................128
5.2 HIGH RISK AND VULNERABLE ROAD USERS ............................................................129
5.2.1 Overview....................................................................................................129
5.2.2 Younger drivers..........................................................................................129
5.2.3 Older drivers.................................................................................................132
5.2.4 Heavy vehicle drivers...................................................................................140
5.3 QUALITATIVE RESEARCH WITH TARGET GROUPS ....................................................148
5.3.1 Objectives...................................................................................................148
xiii
5.3.2 Method .......................................................................................................148
5.3.3 Results ........................................................................................................152
5.3.4 Discussion ..................................................................................................170
5.4 QUANTITATIVE RESEARCH WITH TRAIN DRIVERS AND EXPERTS .............................173
5.4.1 Objectives...................................................................................................173
5.4.2 Method .......................................................................................................173
5.4.3 Results ........................................................................................................175
5.4.4 Discussion ..................................................................................................185
5.5 FRAMEWORK FOR INTERVENTION DEVELOPMENT ..................................................186
5.5.1 Overview....................................................................................................186
5.5.2 Social marketing ........................................................................................187
5.5.3 Intervention mapping .................................................................................188
5.6 SUMMARY..............................................................................................................190
CHAPTER SIX: THE PRESENT CONTEXT OF MOTORIST BEHAVIOUR AT
LEVEL CROSSINGS ..................................................................................................191
6.1 INTRODUCTION ......................................................................................................192
6.2 STUDY AIMS AND RESEARCH QUESTIONS...............................................................192
6.3 INTERVENTION DEVELOPMENT AND IMPLEMENTATION ..........................................194
6.3.1 Overview....................................................................................................194
6.3.2 Heavy vehicle drivers.................................................................................195
6.3.3 Older drivers ..............................................................................................196
6.3.4 Younger drivers..........................................................................................196
6.4 METHOD.................................................................................................................197
6.4.1 Research design..........................................................................................197
6.4.2 Questionnaire measures .............................................................................198
6.4.3 Pilot testing ................................................................................................206
6.4.4 Ethical considerations ................................................................................207
6.4.5 Sample size calculations ............................................................................207
6.4.6 Recruitment strategy ..................................................................................208
6.4.7 Sampling method .......................................................................................209
6.4.8 Procedure and response rate.......................................................................210
xiv
6.4.9 Data management.......................................................................................213
6.4.10 Dependent variables ...................................................................................214
6.4.11 Data analysis ..............................................................................................214
6.5 RESULTS.................................................................................................................215
6.5.1 Heavy vehicle drivers.................................................................................215
6.5.2 Older drivers ..............................................................................................231
6.5.3 Younger drivers..........................................................................................250
6.6 STUDY LIMITATIONS ..............................................................................................268
6.6.1 Widespread limitations ..............................................................................268
6.6.2 Heavy vehicle driver sample limitations....................................................269
6.6.3 Older driver sample limitations..................................................................271
6.6.4 Younger driver sample limitations.............................................................271
6.7 SUMMARY..............................................................................................................272
CHAPTER SEVEN: DISCUSSION ..........................................................................277
7.1 INTRODUCTION ......................................................................................................278
7.2 OVERVIEW OF THE STUDIES, METHODOLOGY AND KEY FINDINGS...........................280
7.2.1 High risk and vulnerable road users at level crossings ..............................281
7.2.2 Planning and development of interventions for each road user group.......285
7.2.3 The present context of motorist behaviour.................................................291
7.3 IMPLICATIONS FOR LEVEL CROSSING SAFETY.........................................................296
7.4 STRENGTHS AND WEAKNESSES OF THE RESEARCH .................................................300
7.5 SUGGESTIONS FOR FUTURE RESEARCH ...................................................................302
7.6 RECOMMENDATIONS TO INDUSTRY ........................................................................303
xv
LIST OF FIGURES
Figure 1: Components of railway level crossing safety ............................................ 16
Figure 2: Estimated cost of collisions at level crossings........................................... 33
Figure 3: Estimated cost of collisions with different types of control ...................... 33
Figure 4: Example of types of rumble strips............................................................. 39
Figure 5: Contributing factors in highway-railway grade crossing collisions. .......... 76
Figure 6: Perceived risk ratings by Canadian drivers ................................................ 80
Figure 7: Driving behaviour responses to potential hazards...................................... 83
Figure 8: Integrated Model of Behaviour Change ................................................... 104
Figure 9: Risk of driver being involved in a casualty road crash............................. 130
Figure 10: Projected older driver fatalities in Australia, 1995-2005. ...................... 134
Figure 11: Austroads class 3 vehicles degree of speeding....................................... 143
xvi
LIST OF TABLES
Table 1: Number of public level crossings by protection type .................................. 20
Table 2: Level crossing occurrences .......................................................................... 27
Table 3: Level crossing accident fatalities ................................................................ 28
Table 4: Car occupants killed due to being hit by a train at a level crossing............. 28
Table 5: Source of rail accident costs in Australia, 1999 ($ million) ........................ 32
Table 6: Estimated collisions with different types of protection .............................. 34
Table 7: Comparison of fatal collision rates ............................................................. 36
Table 8: Fatal road crashes and fatalities, Australia 1981 to 1998 .......................... 142
Table 9: Data collected during formative research phase ........................................ 152
Table 10: Data collected from truck drivers during formative research phase ........ 166
Table 11: Level of aggregate train driver assigned risk to motorist behaviours...... 177
Table 12: Level of regional train driver assigned risk to motorist behaviours ........ 179
Table 13: Level of urban train driver assigned risk to motorist behaviours ............ 180
Table 14: Comparison of means between urban and regional train drivers............. 181
Table 15: Level of expert assigned risk to motorist behaviours .............................. 183
Table 16: Comparison of means between train drivers and experts ........................ 184
Table 17: Items and scales included in the T1 questionnaire................................... 199
Table 18: Items and scales included in the T2 questionnaire................................... 200
Table 19: Structure of the Rural, Remote and Metropolitan Classifications .......... 201
Table 20: Response rates compared across road user groups .................................. 213
Table 21: Company attrition .................................................................................... 217
Table 22: Mean age between companies ................................................................. 217
Table 23: Mean overall scores between shift work and driving at level crossings.. 218
Table 24: Crash involvement between companies................................................... 219
Table 25: Road crashes during past 3 years ............................................................. 219
Table 26: Contributing factors to road crash............................................................ 220
Table 27: Knowledge of level crossing rules and facts............................................ 220
Table 28: Modified DBQ ......................................................................................... 221
Table 29: Self-reported driving behaviour at crossings ........................................... 221
xvii
Table 30: Intended driving behaviour at crossings .................................................. 222
Table 31: Bivariate correlations between dependent and independent variables ... 223
Table 32: Attitudes towards driving at level crossings ............................................ 224
Table 33: Perceived behavioural control whilst driving at crossings ...................... 224
Table 34: Mean scores for subjective norms of others ............................................ 226
Table 35: Belief of the likelihood of being involved in a level crossing collision . 226
Table 36: Beliefs of design and environmental factors at level crossings .............. 227
Table 37: Exposure to level crossing driving........................................................... 228
Table 38: Mean scores for familiarity at level crossings (boom gates) .................. 229
Table 39: Mean scores for familiarity at level crossings (flashing lights).............. 229
Table 40: Mean scores for familiarity at level crossings (passive signs only)........ 230
Table 41: Participant group and area classification at both time points .................. 231
Table 42: Licence conditions as reported by participants........................................ 232
Table 43: Medical conditions suffered by participants ............................................ 233
Table 44: Driving ability.......................................................................................... 234
Table 45: Driving self-assessment ratings between groups ..................................... 234
Table 46: Comparison of crash involvement and rurality at Time 1 ....................... 235
Table 47: Road crashes during past 3 years ............................................................. 236
Table 48: Self-report contributing factors to road crash .......................................... 236
Table 49: Mean overall scores between road crashes and driving instruments ...... 237
Table 50: Bivariate correlations for older drivers ................................................... 239
Table 51: Hierarchical regression of constructs on intention at level crossings ...... 240
Table 52: Attitudes towards driving at level crossings ............................................ 241
Table 53: Perceived behavioural control whilst driving at level crossings.............. 241
Table 54: Subjective norms of others ...................................................................... 242
Table 55: Beliefs of environmental factors at level crossings ................................ 243
Table 56: Exposure to level crossing driving........................................................... 244
Table 57: Familiarity and area classification at Time 1........................................... 244
Table 58: Familiarity with driving at level crossings (boom gates) ....................... 245
Table 59: Familiarity with driving at level crossings (flashing lights) ................... 246
Table 60: Familiarity of driving at level crossings (passive signs only)................. 246
xviii
Table 61: Comparison of experimental group outcome variables ........................... 248
Table 62: Repeated measure analysis of variance for the outcome variables.......... 249
Table 63: Participant group and area classification at both time points .................. 250
Table 64: Education level of younger participants .................................................. 251
Table 65: Mean overall scores between licence type and outcome variables......... 251
Table 66: Mean overall scores between gender and outcome variables ................. 252
Table 67: Comparison of crash involvement and rurality at Time 1 ....................... 253
Table 68: Mean overall scores between road crashes and outcome variables ........ 254
Table 69: Bivariate correlations between dependent and independent variables ... 255
Table 70: Hierarchical regression of constructs on intention at level crossings ...... 257
Table 71: Attitudes towards driving at level crossings ............................................ 258
Table 72: Perceived behavioural control whilst driving at level crossings.............. 259
Table 73: Mean scores for subjective norms of others ............................................ 259
Table 74: Beliefs of environmental constraints whilst driving at level crossings... 260
Table 75: Exposure to level crossing driving........................................................... 261
Table 76: Familiarity and area classification at Time 1........................................... 261
Table 77: Familiarity with driving at level crossings (boom gates) ....................... 262
Table 78: Familiarity with driving at level crossings (flashing lights) ................... 263
Table 79: Familiarity with driving at level crossings (passive signs only)............. 264
Table 80: Comparison of experimental group instrument mean scores................... 265
Table 81: Repeated measure analysis of variance for the outcome variables.......... 267
xix
LIST OF APPENDICES
Appendix 1: Review of models/theories ………………………………….. 364
Appendix 2: Modified Delphi Technique (First Questionnaire) ………….. 386
Appendix 3: Modified Delphi Technique (Second Questionnaire) ………. 390
Appendix 4: Survey instrument used in Study Two ……………………… 397
Appendix 5: Support letter from RACQ inviting Younger Drivers to
participate ……………………………………………………
399
Appendix 6: Support letter from RACQ inviting Older Drivers to
participate ……………………………………………………
401
Appendix 7: Intervention and Control Radio Script for Each Road User
Group ………………………………………………………...
403
Appendix 8: Questionnaires used in Study 3 ……………………………... 412
xx
GLOSSARY OF ACRONYMS AND TERMS
ALCAM Australian Level Crossing Assessment Model
ARA Australasian Railways Association
ARRB Australian Road Research Board
ARSCIG Australian Railway Crossing Strategy Implementation Group
ARTC Australian Rail Track Corporation
ATC Australian Transport Council
ATSB Australian Transport Safety Bureau
BAC Blood Alcohol Concentration
DBQ Driver Behaviour Questionnaire
DIPNR Department of Infrastructure, Planning and Natural Resources
ESCAP Economic and Social Commission for Asia and the Pacific
GDP Gross Domestic Product
HBM Health Belief Model
HIV Human Immunodeficiency Virus
HSE Health and Safety Executive
IM Integrated Model of Behaviour Change
IGA Intergovernmental Agreement
ITS Intelligent Transport Systems
ITSRR Independent Transport Safety and Reliability Regulator
LGA Local Government Association
MoT Ministry of Transport
MUARC Monash University Accident Research Centre
NHTSA National Highway Traffic Safety Administration
NRTC National Road Transport Commission
NTSB National Transportation Safety Board
ODSAQ Older Driver Self Assessment Questionnaire
OECD Organisation for Economic Co-Operation and Development
QR Queensland Rail
QT Queensland Transport
xxi
RACQ Royal Automobile Club of Queensland
RIC Rail Infrastructure Corporation
RLX Railway Level Crossing
SA Shire Association
RTA N.S.W. Roads and Traffic Authority
STD Sexually Transmitted Disease
THC Tetrahydrocannabinol (cannabis)
U.S.A. United States of America
xxii
STATEMENT OF ORIGINAL AUTHORSHIP
The work contained in this thesis has not been previously submitted for a degree or
diploma at any other higher education institution. To the best of my knowledge and
belief, the thesis contains no material previously published or written by another person
except where due reference is made.
Signed: _____________________________________________________________
Date: _______________________________
xxiii
ACKNOWLEDGMENTS
There is of course many people to thank for assisting and supporting my journey
throughout this PhD. Firstly, I would like to thank my husband Matt for his endless
emotional support, particularly when it all seemed too hard. Without your support, I
honestly don’t think I could have sustained my motivation. Secondly, I would like to
thank my parents for their ongoing support of my continuous journey of study. The fact
that they are proud of me doing my PhD has been a significant factor in keeping me
going.
Of course I need to thank my primary supervisor Associate Professor Jeremy
Davey. Thank you for providing me with so many wonderful opportunities to travel
abroad to present at conferences, to conduct media interviews and establish important
contacts. You have helped me keep my sense of humour along the journey, which has
been a saving grace. Also to my associate supervisors Adjunct Professor Vic Siskind
and Associate Professor Barry Watson, thank you for guiding me in the process of
‘working through it all’. Additional to my supervisory team, I need to thank Dr. James
Freeman for helping me clarify things in the end and assisting me in putting all the
pieces in the puzzle together.
Finally, I would like to thank the Rail CRC for their financial assistance and the
rail industry for supporting my journey of inquiry. I received nothing but support from
rail authorities throughout Australia, and hope that this PhD provides ‘real-world’
recommendations for improving level crossing safety for the future.
1
CHAPTER ONE: INTRODUCTION TO THE THESIS
1.1 Introduction ……………………………………………………….. 2
1.2 The research area ……………………………………...................... 3
1.3 Rationale for the research …………………………………………. 4
1.4 Theoretical framework ………………………………..................... 5
1.5 Research objectives ……………………………………………….. 6
1.6 Scope of research ……………………………………..................... 8
1.7 Thesis outline ……………………………………………………... 8
1.8 Summary ………………………………………………………….. 9
2
1.1 INTRODUCTION
Railway level crossing collisions in Australia are a major cause of concern for
both rail and road authorities. Despite the fact that the number of railway collision
fatalities in Australia has fallen in recent years, level crossing collisions constitute a
significant proportion of the national rail toll. Although rail transport is presently one of
the safest forms of land transport, collisions at level crossings are three times more likely
to involve fatalities as compared to all other types of road crashes (Afxentis, 1994). Of
all the types of road crashes which occur, those that involve a collision between a motor
vehicle and train are amongst the most severe. Conservative estimates of level crossing
collisions involving motor vehicles are valued to be approximately $10 million per year
(Bureau of Transport and Regional Economics, 2002), however more recent estimates
from industry experts indicate that this figure is closer to $50 million per year. With
many level crossing fatalities and injuries resulting in coronial inquests, litigation and
negative media publicity, the actions of rail and road infrastructure providers and the
behaviour of motorists, pedestrians and rail users, come under close scrutiny.
Historically, research in this area has been plagued by the rail/road interface and the
separation of responsibilities between rail and road authorities reflecting the social and
political context in which they are contained. With the recent rail reform in Australia
and the public demand for safety at level crossings, motorist behaviour has become a
key priority area.
Accordingly, there is a need to better understand the scope and nature of motorist
behaviour at level crossings, in order to examine the key determinants of behaviour and
associated constructs (e.g. intentions, attitudes, norms, self-efficacy or perceived
behaviour control, perceived risks, skills and abilities of motorists, and perceived
environmental constraints). Additionally, there is great need to investigate whether
targeted educational interventions produce any statistically significant changes in such
constructs in order to develop future road safety countermeasures for both intentional
and unintentional unsafe driving behaviour.
This thesis documents a program of research undertaken for this purpose.
3
1.2 THE RESEARCH AREA
In Australia, at-grade road/railway crossings are commonly referred to as ‘level
crossings’ and there are a large number of these across the rail network. ‘Active’ and
‘passive’ crossings are the two major types of railway level crossings. There are
approximately 9400 public level crossings of which approximately 2650 (28%) have
active protection, 6060 (64%) have passive protection and the remainder have other
control or protection (Ford, 2002). Passive crossings have no dynamic devices to warn
drivers of an approaching train and feature a static array of signs that remain constant.
These crossings have no lights, bells, booms, gates or other active devices warning of an
approaching train (Wigglesworth, 2001). Alternatively the term ‘active’ crossing
encompasses crossings that display a range of dynamic devices to alert motorists of
upcoming danger and impose a requirement to stop in unsafe conditions. Active crossing
devices include features such as flashing lights, bells, boom barriers and wooden gates,
which are activated through track circuitry operated by an approaching train
(Wigglesworth, 2001).
State and Territory rail safety regulators have agreed to use the following
definition of a railway occurrence:
Any accident or incident involving a railway train or other railway vehicles
operated on rails, whether in motion or not, or other event on railway property
affecting the safety of persons and property.
Includes:
• Collision, derailment, fire, explosion, act of God, or other event; and
• Slips, trips and falls on trains or railway infrastructure
Excludes:
• Occurrences in repair shops, not involving a train in motion; and
• Assaults (for national reporting only. Individual State/Territory Rail Safety
Regulators may require assaults to be reported. The later definitions of assaults
are designed to facilitate consistency between those States/Territories)
4
Note: The classification of an occurrence by type (collision, derailment etc) is
determined by the ‘top event’ in the sequence (i.e. the event with the greatest
outcome). This may not necessarily be the final event in a chain of events.
(Australian Transport Safety Bureau, 2004, pxv)
1.3 RATIONALE FOR THE RESEARCH
This program of research was motivated by a strong desire from within the rail
industry to investigate motorist behaviour at level crossing safety to support current
engineering efforts. However, a number of obstacles have hindered research into the
area of level crossing safety. Firstly, level crossings are amongst the most multifaceted
of road safety issues due to the accumulation of rail infrastructure, trains and train
operations (Australian Transport Council, 2003).
Secondly, minimal research has been conducted into collision involvement
patterns of drivers at level crossings, both within Australia and internationally. While
some studies have been conducted in this area, they have tended to be general in
approach and have not specifically examined contributing factors to collision
involvement. As with many crashes, the contributing causes and factors are often
difficult to determine, however a recent investigation of fatal collision at level crossings
supports the notion that human fault is a major contributor (Australian Transport Safety
Bureau, 2002b).
Thirdly, there is a lack of reliable data available relating to the behavioural
characteristics and perceptions of motorists at level crossings. Studies that do exist have
lacked a strong theoretical base to guide the interpretation of results. With many level
crossing fatalities and injuries resulting in coronial inquests, litigation and negative
media publicity, the actions of rail and road infrastructure providers and the behaviour of
motorists, pedestrians and rail users, come under close scrutiny.
Fourthly, there is a need to better understand the scope and nature of motorist
behaviour at level crossings, in order to examine the key determinants of behaviour and
associated constructs (e.g. intentions, attitudes, norms, self-efficacy or perceived
behaviour control, perceived risks, skills and abilities of motorists, and perceived
5
environmental constraints). To date, there has been no research conducted either in
Australia or overseas that has measured changes in behavioural constructs from targeted
educational interventions for specific road user groups. Such investigations assist in
directing both human efforts and financial resources in developing suitable road safety
countermeasures in attempting to change both intentional and unintentional unsafe
driving behaviour at level crossings.
Although engineering approaches have traditionally dominated risk management
strategies in rail safety, research suggests that such technologies are reaching their point
of diminishing returns. Due to the lack of financial viability of continuing to approach
risk management purely from an engineering perspective, the merits of behavioural
research need to be examined for suitability. In Australia, there has been considerable
recognition regarding the importance of human factor approaches to level crossing
safety (Australian Transport Council, 2003). However, little attempt has been made by
authorities to scientifically examine the present context of motorist behaviour at level
crossings or measure changes in behavioural constructs after exposure to educational
interventions. With the spate of several heavy vehicle-train collisions observed in rural
Victoria during 2007, both the public and industry alike are seeking both explanations
and solutions to improve level crossing safety for both road and rail users.
1.4 THEORETICAL FRAMEWORK
This research program has adopted a multidisciplinary approach to theory,
reviewing perspectives from psychology, sociology and public health to explore motorist
behaviour at level crossings. The examination of a broad approach of perspectives is
necessary due to the variety of behaviours involved in collisions and near-misses at level
crossings. The motivation underlying motorist behaviour determines to a large extent
how successful behaviour change strategies (e.g. educational interventions) may be.
Although no one concept is more precise than the other, some are certainly more
useful (Stafford, 1993). Fishbein’s Integrated Model of Behavioural Change (IM) based
largely on the health belief model, theory of reasoned action and theory of planned
behaviour (Fishbein, 2000), appears to be the most useful in exploring motorist
behaviour at level crossings. The three major determinants of intentions measured both
6
directly and indirectly in this model are attitudes, norms and self-efficacy. What is
likely to make this model more suitable in exploring level crossing safety is its
incorporation of both environmental factors and the skills and/or abilities of drivers –
two very important aspects when considering the complex environment that drivers face
at many level crossings. Additionally, the IM includes other relevant predictors of
intention and behaviour such as demographics, personality characteristics, which are
assumed to affect intentions indirectly through influencing underlying beliefs (Fishbein
et al, 2003). This model appears to also be valuable in guiding the development of data
collection instruments. Other factors such as sensation seeking and habitual behaviours
are not systematic psychological theories per se, and are not detected using this IM.
Therefore, such factors will be appraised using conceptual frameworks for explaining
motivations of motorists.
While the IM proposed by Fishbein assists in the exploration of the entire program
of research, it is particularly relevant to Study Three. Study Three has two main
purposes. Firstly, it assists in the planning, development and measurement of change
associated with exposure to targeted educational interventions specific to each of the
three road user groups identified in Study One. Secondly, it examines the present
context of motorist behaviour at level crossings using key constructs from the IM. As
no known research has been conducted that utilises any psychosocial model to explain or
predict level crossing behavior within different road user groups, this research program
uses this model as an exploratory tool rather than a tool to asses the model’s capacity in
explaining such behaviour.
Chapter Three will specifically review Fishbein’s model, while an extensive
review of a variety of theoretical perspectives that have been used to explain motorist
behaviour such as the theory of planned behaviour, health belief model, deterrence
theory and social cognitive theory, are included as Appendix 1.
1.5 RESEARCH QUESTIONS AND OBJECTIVES
The first study in the research program will examine the following important
research questions:
• What types of motorists are at risk of a collision?;
7
• What are the behaviours that motorists exhibit that increase their chances of a
collision?;
• How frequently do incidents and collisions occur?; and
• How frequently are incidents recorded by train drivers?
The overall objective of the second study is to provide information and
understanding about the knowledge, attitudes, norms, beliefs and risk perceptions of
motorists in relation to driving at level crossings. Specific research questions for Study
Two include:
• Do motorists perceive that they are at risk of being involved in a level crossing
collision?;
• What level of knowledge of the road rules exists?;
• What source and medium is believed to be the most appropriate for road users?;
and
• What differences exist between train drivers and experts opinions in terms of
high risk behaviours at level crossings?
Study Three will involve three parts. The aim of Part One of this study is to
develop targeted interventions specific to each of the three road user groups in
accordance with Fishbein’s theoretical model (Integrated Model of Behaviour Change).
Part Two will involve the investigation of the present context of unsafe driving
behaviour at level crossings. This includes examining knowledge, attitudes, beliefs,
perceptions, self-reported and intended behaviour, and environmental constraints of each
road user group. This second part will also involve the examination of the present
context of motorist behaviour at level crossings using key constructs from Fishbein’s
Integrated Model of Behaviour Change. Part Three involves trialing a pilot road safety
radio advertisement using an intervention and control methodology. This part will
investigate the changes in pre and post-test constructs such as intentions, attitudes,
norms, self-efficacy or perceived behaviour control, perceived risks, and perceived
environment constraints whilst driving at level crossings.
8
1.6 SCOPE OF RESEARCH
This thesis examines the behaviours of motor vehicle drivers at railway level
crossings. Collisions within station precincts, those at non-crossing locations,
trespassing, suicides and those in freight terminals are not included in this thesis.
1.7 THESIS OUTLINE
This research program is the first of its kind in examining the present context of
motorist behaviour at level crossings using an integrated model of behaviour change.
Additionally, it is the first of its kind in measuring change in key behavioural constructs
after exposure to targeted educational interventions. Although other ‘educational’
campaigns exist in this field, no interventions have been guided by empirical research or
psychological theory. The structure of the thesis reflects the specific tasks undertaken as
part of the research.
Chapter Two: Reviews the available evidence relating to collisions and near-
misses at level crossings, drawing on a variety of data sources. The major issues
examined are: the prevalence of collisions at level crossings; the behaviour of road
users; the personal, social and environmental factors contributing to the behaviour; and
the effectiveness of current countermeasures to manage risk. At the conclusion of the
chapter, a number of key research questions are identified to guide the research.
Chapter Three: Reviews Fishbein’s integrated model to examine the present
context of motorist behaviour at level crossings and assist in the planning, development
and measurement of change associated with exposure to targeted educational
interventions.
Chapter Four: Documents the first study undertaken as part of the research. This
study involved two types of data collection methods: (1) a modified Delphi technique to
gather information from road and rail experts to gain an informed judgment, and (2)
focus groups with train drivers to provide insight into the risk taking behaviour of
motorists at level crossings.
Chapters Five: Documents the second study undertaken as part of the research. It
involved formative research with members of the three road user groups identified in the
first study, as part of the planning and development targeted educational interventions
9
for the third study. Two forms of formative research (qualitative and quantitative) were
used to provide methodological triangulation and ensure reliability of the data.
Chapter Six: Reports on the findings from the third study which involved an
exploration of the present context of motorist behaviour at level crossings (using key
constructs from Fishbein’s IM). Additionally, this third study involved the planning and
development of targeted educational interventions, and the measurement of change
associated with intervention exposure. Older (60 plus years), younger (17-24 years) and
heavy vehicle driver groups were recruited through various agencies. Pre and post-test
questionnaires were provided to participants. The majority of analyses conducted with
heavy vehicle drivers were of an exploratory nature due to small sample sizes.
However, investigations of older and younger drivers also involved parametric
inferential statistics.
Chapter Seven: Concludes the thesis by discussing the findings from all three
studies in the context of the literature reviewed, as well as identifying limitations of the
research undertaken. Additionally, it presents the implications of the research findings
for further research and provides recommendations to industry stakeholders.
1.8 SUMMARY
This chapter has provided a brief overview of some of the obstacles that have
hindered research into the area of level crossing safety, as well as discussing some of the
theoretical issues relevant to understanding motorist behaviour. To date, limited research
has been conducted into motorist behaviour at level crossings, both within Australia and
internationally. Furthermore, financial resources have not customarily been directed
towards researching motorist behaviour at level crossings as fatalities constitutes less
than 1% of the national road toll. However, with the recent spate of several heavy
vehicle-train collisions observed in rural Victoria and the media attention these
collisions have attracted, both the public and industry alike are seeking explanations and
solutions to improve level crossing safety for both road and rail users alike.
The foundations for the current research program will be laid in Chapter Two
which will review the research literature relevant to the study of motorist behaviour at
level crossings. The remaining chapters will present and discuss the findings of three
10
specific studies undertaken to improve the existing body of evidence relating to level
crossing motorist behaviour.
11
CHAPTER TWO: LITERATURE REVIEW
2.1 Introduction ………………………………………………………... 12
2.2 Australia’s rail industry ……………………………………………. 12
2.3 Level crossing protection systems …………………………………. 18
2.4 Level crossing collision statistics in Australia …………………….. 25
2.5 Engineering measures ……………………………………………… 36
2.6 Laws and enforcement ……………………………………………... 56
2.7 Education, publicity and training ………………………………….. 60
2.8 Human factors contributing to collisions ………………………….. 72
2.9 Research questions ………………………………………………… 97
2.10 Summary …………………………………………………………… 98
12
2.1 INTRODUCTION
The following chapter will review the available literature relating to the nature and
extent of level crossing collisions and the effectiveness of relevant engineering,
enforcement and educational countermeasures. Firstly, an overview of the railway
industry in Australia will be explored to gain a greater understanding of the context in
which level crossing safety exists. This review of engineering measures focuses on
presenting current and potential risk management solutions for protecting motorists at
level crossings. Given the limited scientific evidence that is available that evaluates
such engineering measures on influencing motorist behaviour, the review is limited in its
critical interrogation of the literature.
Among the key questions that are explored in this literature review are: how
prevalent are collisions at level crossings; what are the contributing human factors to
collisions at level crossings; what engineering, enforcement and educational
countermeasures are currently in place to reduce collisions at level crossings; and how
different road user groups are at risk of being involved in a collision. These questions
are informed by the examination of human factors contributing to collisions that makes
up a substantial portion of this chapter. In this examination a more critical interrogation
of the literature is presented for educational programs that aim to influence motorist
behaviour at level crossings.
The aim of the chapter will be to consolidate the available research evidence and
identifying gaps in current knowledge relating to motorist behaviour at level crossings.
By reviewing the literature, it will lay a foundation for the subsequent program of
research that is undertaken for this thesis.
2.2 AUSTRALIA’S RAIL INDUSTRY
2.2.1 Economic importance
Australia’s railway industry is critical to the nation’s economy (Hill, 2001, p101).
The 40,000 kilometre rail network “serves intercapital markets and many important
economic regions” (Hill, 2001, p1). The majority of companies in the industry are
private profitable ventures trading in highly competitive domestic and international
13
markets (Australian Rail Association Incorporated, 2002) and generate 1.6% of
Australia’s gross domestic product (GDP) with output of goods and services worth $8
billion each year, including exports worth $0.5 billion per year (excluding coal and iron
ore) (Australian Rail Association Incorporated, 2002).
2.2.2 Rail activities
The rail industry in Australia is very diverse, consisting of rail operators (freight,
passenger, tourist and heritage), manufacturers, suppliers, consultants, track access
corporations, maintenance and construction contractors, logistics providers and a wide
range of other companies covering all sectors of the industry (Australian Rail
Association Incorporated, 2002). Rail freight services are a fundamental part of the
distribution process for interstate freight and a range of regional produce and bulk export
commodities (Hill, 2001), while urban rail systems are an essential part of the transport
infrastructure of cities, providing enormous economic, social and environmental benefits
(Australian Rail Association Incorporated, 2004b). According to the Bureau of
Transport and Regional Economics (2002), rail has consistently carried around a third of
Australia’s domestic freight in terms of tonne-kilometres during the past 25 years,
although increased competition from road transport and lack of investment in rail
infrastructure has led to a steady decline in the proportion of non-bulk interstate freight.
Around 80% of rail transport activity in Australia, measured in terms of the distance
traveled by both freight and passenger trains, occurs in the eastern jurisdictions of New
South Wales, Victoria and Queensland (Australian Transport Safety Bureau, 2003).
2.2.3 Reform agenda
Rail transport in Australia has undergone significant reform since the early 1990’s,
triggered by competition from road transport and government activity (Bureau of
Transport and Regional Economics, 2002). In November 1996, the Federal government
announced its rail reform agenda to provide the impetus to make rail competitive and
thus ensure it continues to be an integral part of Australia’s national transport system
14
(Department of Transport and Regional Services, 2004). The most significant outcome
from this rail reform process has been the fragmentation of the industry into multiple
operational, management and regulatory jurisdictions (Department of Transport and
Regional Services, 2004). As the prominence of interstate rail operations increased and
the rail industry was opened up to private operators, pressure was mounting to eliminate
State-based obstacles (such as differing safety standards and accreditation, to efficient
interstate operations). Now, the rail industry is progressively more seen as a national
industry not as a series of State-based industries (Australian Rail Association
Incorporated, 2004a). The Federal government believed that this change was necessary,
with Federal, State and Territory transport ministers signing an ‘Intergovernmental
Agreement on Rail Safety’, which came into effect in May 1996. This
Intergovernmental Agreement (IGA) is both a recognition of and commitment to the
need for a nationally consistent approach to rail safety (Department of Transport and
Regional Services, 2004). Safety arrangements under the 1996 IGA have reached a stage
of maturity whereby the administrative process is now functioning and safety is
considered a high priority (Australian Transport Council, 2002).
2.2.4 Rail safety management
Australia’s rail industry has been changing rapidly as a result of initiatives by
Commonwealth, State and Territory governments to reform historic structures,
policies and practices, and by rail organisations to improve customer service,
safety, asset quality and commercial performance.
(Affleck Consulting Pty Ltd, 2003, pv)
However, according to the Australian Transport Council, the safety arrangements
in Australia “continues to expose shortcomings as the rail industry develops and in the
distinction between operators on the interstate and intrastate networks become
increasingly redundant” (Australian Transport Council, 2002, p3). These shortcomings
relate mainly to interface management, performance and risk information, role
differentiation and administration (Australian Transport Council, 2002). Interface
15
management between the rail and road sectors requires specific attention to safety
management processes. The emergence of level crossing safety as a key priority area is
the result of the changing face of Australia’s railway industry.
Although safety at level crossings is only one aspect of transport safety within the
whole transport system (Australian Transport Council, 2003), it constitutes an
identifiable proportion of the rail toll. This no doubt has been the impetus for the release
of the ‘National Railway Level Crossing Safety Strategy’ by the Australian Transport
Council (ATC) in August 2003. The objective of this strategy is to reduce the “number,
cost and trauma of crashes between trains and any road users by the most cost-effective
means” (Australian Transport Council, 2003, p3). Additionally, level crossing safety has
for the first time been included in the National Road Safety Action Plan 2003/2004 but
remains excluded from most State and Local government road safety strategies. With
level crossing safety currently high on the rail industry agenda and the growth in level
crossing safety activities occurring throughout Australia, research examining key
determinants of motorist behaviour and road safety countermeasures is timely. Figure 1
illustrates the multifaceted system in which level crossing safety is situated (Australian
Transport Council, 2003).
16
Figure 1: Components of railway level crossing safety
2.2.5 National strategy initiative
Prior to 2002, there had been considerable activity with respect to level crossing
safety in all parts of Australia, but without much consistency or coordination (Wallace et
al., 2006). Important initiatives were being undertaken by State level crossing
committees, road agencies and the rail industry but they were largely unsystematic,
independent and unrelated. Several issues of critical local importance arose and were
being addressed in individual jurisdictions, when government Transport Ministers at the
ATC requested a framework for national level crossing safety to be prepared. In 2003,
the ATC endorsed the National Railway Level Crossing Safety Strategy, together with a
draft action plan for specific initiatives. The Australian Railway Crossing Strategy
Implementation Group (ARCSIG) was formed to coordinate the wide range of activities
on behalf of the ATC via the governments' National Transport Coordinating Committee,
the Standing Committee on Transport and the Rail Group.
Regulation
Industry practiceInformation
FundingEnforcement
Governmentpolicy
Technology Communityresponse
RoadInfrastructure
Road Vehicles
RailInfrastructure
Train Drivers
Rail Vehicles
Road Users
Environment
Education
17
The national strategy identified that there were many government activities that
required some measure of improvement. Consequently, the national strategy broadly
covers education, engineering, enforcement and policy issues. Upgrades to protection
systems at specific sites are outside the scope of this strategy as the responsibility lies
with government and the rail sector. Each national strategy activity is owned by
individual agencies and ARCSIG distributes information and provides liaison. Several of
the actions have been completed and several more have been added to the program.
ARCSIG is active in further improving the coordination between policy, engineering,
road and rail, education and enforcement agencies including state and national level
crossing committees.
2.2.6 Australasian railway level crossing behavioural strategy
After commencement of this project in early 2004, the Australasian Railways
Association (ARA) initiated the development of a behavioural strategy in early 2005.
This strategy seeks to form partnerships between industry, government, and road and rail
stakeholders to develop education and enforcement programs. This strategy is aiming to
introduce ‘Operation Lifesaver’ in each jurisdiction. The merits of ‘Operation Lifesaver’
are discussed in Chapter Two (Literature Review). The objective of this strategy is to
reduce the number, cost and trauma associated with level crossing collision through
behavioural programs that are developed and delivered in all jurisdictions. This is
thought to be achieved by:
• Macro level programs – encompassing mass media programs focusing on radio,
television, billboards and press advertisements;
• Micro level programs – encompassing targeted education and awareness
initiatives (face-to-face education, schools, high risk groups, high risk locations);
and
• Enforcement – encompassing programs aimed at penalising non-compliance of
the road rules at level crossings.
(Australasian Railways Association, 2006)
18
2.2.7 Roles and responsibilities for level crossing safety
Numerous agencies share the responsibility for ensuring safety at level crossings,
with the strategic coordination and oversight occurring through each State and Territory
level crossing safety committee. For example, in New South Wales there are nine main
agencies that share responsibility for level crossings. Their role in this responsibility is
summarised below.
• Australian Rail Track Corporation (ARTC) - manages and maintains the NSW
country and interstate rail network under a 60-year lease from the State
Government. ARTC also maintains the remaining country rail network under
agreement to the Rail Infrastructure Corporation (RIC).
• Department of Infrastructure, Planning and Natural Resources (DIPNR) - drives,
co-ordinates and streamlines land use and transport planning, infrastructure
development and natural resource management in NSW.
• The Independent Transport Safety and Reliability Regulator (ITSRR) - ensure
public transport and commercial railway operations are safe for use by the
communities and businesses of NSW.
• Local Government Association and Shires Association of NSW (LGA & SA) –
are the peak bodies representing the interests of NSW metropolitan, regional and
rural councils to other spheres of government and the wider community.
• Ministry of Transport (MoT) - provide independent, considered policy advice
and financial and strategic coordination for the transport portfolio to improve
passenger and freight transport service outcomes for the people of NSW.
• NSW Police - aims to protect the community and property by, preventing,
detecting and investigating crime, monitoring and promoting road safety,
maintaining social order, performing and coordinating emergency and rescue
operations. Other major services include traffic control, communications,
intelligence analysis, anti-terrorist negotiation and security coordination.
• Rail Infrastructure Corporation (RIC) - owns the NSW country rail network on
behalf of the State Government.
• RailCorp - owns and maintains the rail infrastructure in the greater metropolitan
Sydney region and delivers CityRail and CountryLink passenger services. It was
19
created in 2004 to provide safer, cleaner, more secure and more reliable rail
transport.
• NSW Roads and Traffic Authority (RTA) - is responsible for promoting road
safety and traffic management, driver licensing and vehicle registration. It is also
responsible for the maintenance and development of the national highway and
State road network in NSW. It provides funding assistance to local councils for
regional roads and to a limited extent, for local roads.
(NSW Government, 2006)
2.3 LEVEL CROSSING PROTECTION SYSTEMS
2.3.1 Types of protection systems
In Australia, at-grade road/railway crossings are commonly referred to as ‘level
crossings’ and there are thousands of these across the rail network. Such road/rail
intersections are unique in the transport sector as they present the only case of two
different infrastructures placed under different responsibilities and traveled by vehicles
with dramatically different performances (United Nations, 2000). Level crossing
protection systems used 60 to 100 years ago, such as boom gates designed to prevent
frightened horses from charging forwards into approaching trains as well as red flashing
lights which emulate the red lanterns that railroad employees swung back and forth to
warn of trains, still feature as operating procedures for rail safety (Green, 2002). Despite
the fact the rail infrastructure in Australia has developed substantially over the past
century, the original infrastructure of level crossings has changed very little (McBride,
2002). Level crossings today occur across a variety of road access types (highways,
other public roads, private roads, and access for rail agencies), road user categories
(motor vehicles, pedestrians, agricultural machinery, and other work vehicles such as
plant machinery) and rail use types (main lines, secondary main lines, branch lines,
heavy haulage lines, and restricted lines subject to seasonal use or tourism activities)
(McBride, 2002). Although protection systems operating at level crossings are less
visually effective than those used at highway intersections (Green, 2002), rail authorities
20
do not have the financial resources to update all level crossing protection systems in
Australia (Afxentis, 1994).
‘Active’ and ‘passive’ crossings are the two major types of railway level crossings
in Australia. Active crossings incorporate devices that warn motorists when it is safe to
use the crossing, either by visual or auditory cues (boom gates and/or twin alternating
flashing red lights). Passive crossings do not warn the motorist of the proximity of an
approaching train (stop or give-way signs only). Passive crossings (n=6060, 64%) are
the most common protection system for level crossings in Australia, followed by active
crossings (n=2649, 28%), with the remainder having other control or protection (Ford,
2002). Table 1 below illustrates the types of level crossings in each jurisdiction.
Queensland has the largest number of level crossings across the rail network with 37%
of Australia’s total number of level crossings (n=3500), followed by Victoria with 24%
(n=2209), 16% in Western Australia (n=1518) and 12% in New South Wales (n=1144).
Table 1: Number of public level crossings by protection type
Source: Ford & Matthews (2002)
21
2.3.2 Definitions
2.3.2.1 Passive crossings
Passive crossings have no dynamic devices to warn motorists of an approaching
train and feature a static array of signs that remain constant. These crossings have no
lights, bells, booms, gates or other ‘active’ device warning of an approaching train
(Wigglesworth, 2001). According to the Australian Standard (AS1742.7 – 1993)
‘Manual of uniform traffic devices – Part 7: Railway crossings’ (Standards Australia,
1993), a passive crossing is defined as:
Control of the movement of vehicular or pedestrian traffic across a railway level
crossing by signs and devices, none of which are activated during the approach or
passage of a train, and which rely on the road user detecting the approach or
presence of a train by direct observation.
2.3.2.2 Active crossings
Active crossing devices include features such as flashing lights, bells, boom
barriers and wooden gates, which are activated through track circuitry operated by an
approaching train (Wigglesworth, 2001). Flashing light crossings consist of (at least)
two sets of twin alternating flashing red lights installed to face each road approach
direction and are activated through track circuitry (Wigglesworth, 2001). Boom barrier
crossings are equipped with two booms hinged about a horizontal axis, one on each side
of the crossing, and are supplemented by flashing lights and warning bells which are
also activated through track circuitry operated by an approaching train (Wigglesworth,
2001).
According to the Australian Standard (AS1742.7 – 1993) ‘Manual of uniform
traffic devices – Part 7: Railway crossings’ (Standards Australia, 1993), an active
crossing is defined as:
Control of the movement of vehicular or pedestrian traffic across a railway level
crossing by devices such as flashing light signals, gates or barriers, or a
combination of these, where the device is actuated prior to and during the passage
of a train through the crossing.
22
2.3.3 Australian standards
A project commissioned by Austroads to examine safety at passive railway level
crossings that was timed to coincide with the ‘7th International Symposium on Railroad-
Highway Grade Crossing Research and Safety’, involving a workshop with
representatives from key stakeholder organisations, found that there needed to be some
changes made to the previous Australian Standard (1742.7-1993) for road signs and
marking at both active and passive level crossings.
The project identified four issues that required investigation. Firstly, the issues of
how best to implement a system of crossing identification numbers at both types of
crossings to enable control centres and emergency services to receive and transmit
accurate locations in the event of collisions or technical faults with protection systems
were raised (Cairney et al., 2002b). The second issue was whether it is possible to
develop an advance warning sign that coveys the message that a passive crossing is
approaching. Anecdotal evidence suggests that it is not clear that the distinction
between advance warning of passive and active crossings is intuitively obvious to
Australian motorists. A study by Mitsopoulos et al (2002) found that many road users
appear to be unaware that many level crossings do not have active protection. In this
driving simulator study, the majority of participants believed that all or almost all of the
railway level crossings in Victoria have active protection, although the actual proportion
is less than half (Mitsopoulos, 2002). The third issue was whether it is feasible to
provide a warning that high-speed trains use the railway line. The fourth issue suggested
by the project was whether it is feasible to provide an indication of general level of risk
associated with level crossings (e.g. use advisory speed as a way of indicating the risk to
road users) (Cairney et al., 2002b).
The current Australian Standard (1742.7) ‘Manual of uniform traffic control
devices’ (Part7: Railway crossings) was approved on behalf of the Council of Standards
Australia on the 19th January 2007 (Standards Australia, 2007). The principal changes
and additions to the previous Australian Standard (1742.7-1993) are summarised as
follows:
• The Standard now promotes use of the red background position sign, R6-25, for
new or replacement signs in preference to the open ‘crossbuck’ sign, R6-24;
23
• Provision is made for active advance warning of the activation of railway
crossing signals under certain conditions;
• More detail is given for sight distance requirements at passive control crossings
for stop and give-way sign control;
• The need to avoid unsafe queuing of traffic on railway crossings upstream of
traffic signals is recognised and the use of corrective measures including signs
and box markings are specified; and
• Standards for pedestrian crossing treatments at railway crossings have been
substantially upgraded and now include provision for people with disabilities.
(Standards Australia, 2007, p2)
However, there remain a few safety issues that are not able to be addressed by this
current Standard. Firstly, the length and mass of heavy vehicles is the most prominent
issue with this Standard.
Long and heavy vehicles with slow acceleration from the stopped position can,
when sight distance along the railway is poor, have difficulty starting up and
clearing a crossing before the arrival of a previously unseen train unless there is
some form of active control to warn that a train is coming. Likewise, long vehicles
can present problems if there is an intersection or other vehicular check point
close to the crossing on the departure side and the rear of such a vehicle
inadvertently fails to clear the conflict area.
(Standards Australia, 2007, p5)
Secondly, the Standard indicates that crossings that are infrequently used (such as
those used seasonally or in isolated areas) “present operational and management
challenges that can only be partially met (if at all in some cases) by application of this
Standard” (Standards Australia, 2007, p5). Additionally, the Standard does not provide
any guidance about upgrading crossings to the next treatment (i.e. passive control to
active control or active control to crossing closure), but rather refers readers to the
ALCAM (Australian Level Crossing Assessment Model) model.
24
2.3.4 International comparison of protection systems
2.3.4.1 Developed countries
In the United States in 2001, there were 154,084 public crossings and 98,430
private crossings (Federal Railroad Administration., 2004) intersecting 200,000 miles of
railroad track (Hall, 2004). In the United Kingdom, there are approximately 8000 level
crossings of various types on the UK national rail network, of which approximately 30%
allow only pedestrians to cross the railway (Rail Safety and Standards Board, 2004). Of
the crossings that are used by motor vehicles (n=5600), there are 1675 (34%) active
crossing and 3967 (66%) passive crossings (Rail Safety and Standards Board, 2004).
During the past ten years in the United Kingdom there has been an overall decreasing
trend in the total number of crossings across the rail network. The Rail Safety and
Standards Board (2004) propose that this is a result of the recognition that level
crossings represent a high risk element of the network that needs to be controlled and
lessened. In Finland, at the beginning of 2001 on the rail network of 5854 km there were
5162 level crossings, with 82% (n=4219) having passive protection (Pajunen, 2004).
These three examples illustrate that level crossings occur in large numbers in rail
networks in developed countries.
2.3.4.2 Developing countries
In developing countries, unprotected crossings are very frequent whereby a large
number of crossings are unofficial with construction of the crossing being done by local
individuals without approval from either the railway or road authorities (United Nations,
2000). The Indian Railway network has a route length of 62,495 kilometres with a total
of 40,445 level crossings being an average of one level crossing every 1.5 kilometres
(United Nations, 2000). Of this total number of level crossings in India, nearly 40% have
active protection (manual barriers) while 7% are completely unprotected with neither
passive or active protection systems (United Nations, 2000). In recent years, there has
seen a shift in the Indian Railways policy regarding protection systems of level
crossings. Level crossings with high traffic volumes (either by road or rail) are now
‘manned’ to open gates to road users (United Nations, 2000). Additionally, subject to the
availability of government funds, crossings which have a ‘traffic moment’ (i.e. train
25
movements x motor vehicle movements) of 100, 000 per day or more are being replaced
by either under or over-passes (United Nations, 2000). Vietnam has one of the densest
railway system with level crossings in Asia, with more than three times the level
crossing density of India (United Nations, 2000). Vietnam’s railway system has
approximately 4842 level crossings over a total network length of only 2712 kilometres
(one level crossing every 0.5 kilometres) (United Nations, 2000). Just under 75% (n=3,
600) of level crossings are unofficial crossings (not provided by Vietnam Railways) and
nearly 93% of all level crossings in Vietnam have no protection system (United Nations,
2000). In contrast to other countries throughout the world, Iran has a relatively small
number of level crossings. In Iran, there is a total of 5995 kilometres of railway network
with a total of 418 level crossings (344 official and 74 unofficial) (United Nations,
2000), equating to one crossing every 14.3 kilometres (United Nations, 2000). Of these
total number of crossings, 217 (69.52%) are equipped with active protection (warning
lights and barrier) and the remainder have no form of protection (United Nations, 2000).
2.4 LEVEL CROSSING STATISTICS IN AUSTRALIA
2.4.1 Data recording
Currently, exact statistics of level crossing fatalities in Australia are difficult to
determine. There are a couple of significant reasons for this. Firstly, vehicle-train
collisions are rare events, and therefore it is difficult to obtain valid information about
preceding behaviour (i.e. likelihood of a suicide attempt) and the frequency of near-
misses. Secondly, the ATSB, the chief organisation for transport safety, includes some
level crossing fatalities in the national road toll and some in the national rail toll. When
both a motor vehicle and a train are involved in a fatal collision, the ATSB counts the
fatalities as motor vehicle fatalities rather than level crossing fatalities (following ICD-9)
(Australian Transport Safety Bureau, 2002b). The ATSB term a 'level crossing accident'
as those involving a collision between a railway train or other railway vehicle and a road
vehicle or person on a public street (trams on public streets excluded) (Australian
Transport Safety Bureau, 2002b). A person is counted as a level crossing fatality if the
26
death occurs up to 12 months after the collision. For the purposes of this thesis, only
vehicle-train collisions are being examined.
Thirdly, most jurisdictions in Australia use different methods of categorising and
recording collisions. Until there exists a systematic method for ensuring that all
jurisdictions categorise and record collisions identically, it is difficult to make accurate
comparisons in collision data. Cairney (2002) recommends that the usefulness of
information from collisions at level crossings would be greatly improved by including
the variables that are currently collected for road crashes, such as vehicle type and driver
characteristics. As a consequence of the differing methods of recording of level crossing
collisions in Australia, there is a lack of definitive evidence available relating to the
extent and nature of level crossing collisions. Therefore, for the purposes of the review
of collision and fatality data for this thesis, data from the ATSB will be predominantly
used.
2.4.2 Incidence of occurrences
Due to the large number of near-misses at level crossings that do not involve a
fatality but have the potential to result in a catastrophic event such as a train derailment,
it is important to examine the occurrences of near-misses as well as collisions involving
fatalities. To date, the ATSB is unaware of the exact number of level crossing
occurrence rates in each jurisdiction (Australian Transport Safety Bureau, 2004),
however anecdotal evidence from train drivers suggests that such occurrences are far
more frequent than reported.
A level crossing ‘occurrence’ as provided by the ATSB (2004):
Any collision of a train or rolling stock with either a road vehicle, person, level
crossing safety equipment or gate, or any other occurrence that compromises
safety at a level crossing. Includes: cases of road vehicles causing damage to
gates, barriers or other equipment at level crossings; near miss incidents; any
case of a train running onto a level crossing when not authorised to do so; any
failure of equipment at a level crossing which could endanger users of the road or
path crossing the railway.
27
The latest available data from the ATSB on level crossing occurrences is provided
in Table 2. This data excludes level crossing occurrences involving tramways crossing
roadways and is provided to the ATSB from each jurisdiction in Australia. Data from
2001 and 2002 indicates that Queensland (n=673) has the highest number of occurrences
followed by Victoria (n=543) and New South Wales (n=526).
Table 2: Level crossing occurrences
Source: Australian Transport Safety Bureau (2004) ‘rail occurrences’
2.4.3 Incidence of fatalities
A study by the ATSB focusing on collision details recorded in the ‘fatality crash
database’, a national database holding records of collisions on public roads resulting in
at least one fatality, indicated that in any given year covered by this database, level
crossing collisions constituted no more than one per cent (1%) of fatal road crashes
(Australian Transport Safety Bureau, 2002b). The actual annual figures range from
0.5% to 1.0%, with an average of 0.7% (Australian Transport Safety Bureau, 2002b).
This database covered the years 1988, 1990, 1992, 1994, 1996, 1997 and 1998 (only part
of the year to date, as some of the relevant coroner’s reports have not yet been received)
(Australian Transport Safety Bureau, 2002b). The reason for recordings of these years is
that the database was only updated every second year up to 1996 and annually after this
year. The 87 cases in the study sample of the fatality database are the fatal collisions at
level crossings that occurred in these years. Compared to other road crashes, the number
of fatalities per 100 fatal level crossing collisions was slightly higher than the number of
fatal road crashes (Australian Transport Safety Bureau, 2002b). The ATSB database
indicates that there were 120 fatalities per 100 fatal level crossing collisions compared
with 113 fatalities per 100 other fatal road crashes (Australian Transport Safety Bureau,
2002b).
28
Table 3 shows the total number of deaths in Australia due to level crossing
collisions each year in the period 1997-2002, the period for which data is currently
available from the ATSB. It should be noted that these are fatalities due to collisions
between trains and road vehicles or pedestrians on public streets. Deaths due to
collisions between trains and road vehicles on private roadways are excluded as well as
suicides. A suicide is defined here as a death that a coroner has found to be a suicide.
Table 3: Level crossing accident fatalities
Source: Australian Transport Safety Bureau. (2003b). N.B. This data includes both motor vehicle occupants and pedestrians.
From 1997-2002, there were 74 deaths due to collisions between trains and motor
vehicles at level crossings (see Table 4). These 74 deaths represented less than 1% of the
national road toll in the same period. Nearly 70% of the car occupants who died as a
result of level crossing collisions were males, but no particular age group stands out
(Australian Transport Safety Bureau, 2003). Occupants of pick-up trucks or vans are
included with only three such fatalities seen in this time series (Australian Transport
Safety Bureau, 2003).
Table 4: Car occupants killed due to being hit by a train at a level crossing
Source: Australian Transport Safety Bureau. (2003b).
29
2.4.3.1 Nature of fatalities
According to the ATSB study of 87 fatal collisions at level crossings (Australian
Transport Safety Bureau, 2002b), the point of impact was more often the front of the
train rather than the side of the train. In 66% of this study sample, the point of impact
was at the front of the train, with 16% being at the side of the train and the remaining
being unknown (Australian Transport Safety Bureau, 2002b). The majority of collisions
occurred in a rural area or urban centre away from a capital city (67%), while 18%
occurred in a capital city and in the remainder of cases the location was unknown
(Australian Transport Safety Bureau, 2002b). With the bulk of collisions not occurring
in capital cities, it stands that only ten percent of collisions occurred at crossings with
boom gates. Forty-four (44%) percent of collisions occurred where the protection
system was passive, while 41% occurred where the crossing was some other type of
active warning system (Australian Transport Safety Bureau, 2002b).
Weather conditions were of little significance in contributing to collisions at level
crossings, as indicated in the ATSB fatality database. Eighty-five percent (n=73) of
collisions occurred in fine weather, with 84% of collisions occurring on a dry road
(Australian Transport Safety Bureau, 2002b). Compared to other fatal road crashes,
adverse weather or road conditions accounted for a slightly higher number of fatalities at
level crossings, with 13% (n=11) being due to such conditions compared with 9%
(n=1043) of other fatal road crashes (Australian Transport Safety Bureau, 2002b). In
terms of time of day, 83% of collisions at level crossings occurred in daylight (excluding
dawn or dusk) (Australian Transport Safety Bureau, 2002b). The majority of collisions
occurred on a weekday during the day (63%), while 14% occurred at night, two percent
at dawn and one percent at dusk, with the remainder unknown (Australian Transport
Safety Bureau, 2002b).
Unintended road user error was found to be more common in level crossing
collisions than other fatal road crashes, as indicated by the ATSB database (Australian
Transport Safety Bureau, 2002b). Forty-six percent (n=40) of level crossing collisions,
while only 22% (n=2708) of other fatal road crashes were recorded to be due to
unintended road user error (Australian Transport Safety Bureau, 2002b). The influence
of alcohol and/or drugs was significantly less common with only 9% (n=8) of collisions
30
involving alcohol and/or drugs, compared to 31% (n=3746) of other fatal road crashes
(Australian Transport Safety Bureau, 2002b).
Fatigue was not considered to be a major factor in level crossings collisions, with
3% (n=3) involving fatigue compared with 8% (n=925) of other fatal road crashes
(Australian Transport Safety Bureau, 2002b). Excessive speed was also less likely to be
a major factor, with only 7% (n=6) of level crossing collisions involving speed
compared to 23% (n=2794) of other fatal road crashes (Australian Transport Safety
Bureau, 2002b). ‘Other risk taking’, as recorded by the ATSB fatality database,
accounted for a similar percentage of fatalities, with 3% (n=3) involving risk taking
compared with 5% (n=560) of other fatal road crashes (Australian Transport Safety
Bureau, 2002b). The ATSB does not specify what ‘other risk taking’ is defined as.
2.4.4 Cost of collisions
Level crossing collisions have been shown to result in enormous human and
financial cost to society (Lobb, 2001). The Bureau of Transport and Regional
Economics (2002) provide the following information pertaining to collisions:
Accident costing is an inexact science. Cost estimates depend on the particular
costing approaches used, the number of accident cost components that can be
estimated, the quality and quantity of available data and the value of key
parameters used (such as the discount rate).
(px)
Although the quality of data used in the Bureau of Transport and Regional
Economics was reported as being low compared to the quality of data available for the
costing of aviation accidents and road crashes (Bureau of Transport and Regional
Economics, 2002), the total cost of rail collisions was estimated in 1999 to be $196
million (Bureau of Transport and Regional Economics, 2002). The total for level
crossing collisions was estimated to be approximately $10 million (Bureau of Transport
and Regional Economics, 2002). However, both railway experts and insurance
companies in Australia have revoked this estimation stating that it is grossly
31
conservative. Table 5 below illustrates the source of rail accident costs in Australia in
1999.
In 2003, Peter Cairney, an Australian expert in road safety, estimated the cost of
collisions at level crossings using the casualty and collision data presented by Ford and
Matthews (2002) and estimates of collision costs provided in 2000 by the Bureau of
Transport Economics (BTE) (Bureau of Transport Economics, 2000). The estimates
calculated by Cairney (2003) were based on these two documents, although Cairney
recognises some incompatibilities in the approaches adopted by the authors. This data
on property collisions was extracted from spreadsheets provided by Ford and Matthews
(2002) and the value of a property damage collision developed by the BTE allocated to
these collisions. Cairney (2003) points out that it is unknown how the distribution of
costs arising from collisions with trains compares to the distribution for all traffic
crashes, and hence regards his procedure as an approximation. Cairney (2003) also
extracted the total number of fatalities, serious injuries and minor injuries from Ford and
Matthews (2002) spreadsheets and applied relevant BTE costs for each. Figure 2
(Cairney, 2003) shows these totals for each jurisdiction and each class of protection
system is illustrated in Figure 3 (Cairney, 2003).
32
Table 5: Source of rail accident costs in Australia, 1999 ($ million)
Source: Bureau of Transport and Regional Economics. (2002). N.B. All figures in 1999 dollars, are based on a discount rate of 4 percent and are rounded to the nearest million dollars.
33
Figure 2: Estimated cost of collisions at level crossings
Figure 3: Estimated cost of collisions with different types of control
34
The breakdown of cost by protection system is illustrated in the table below.
Cairney (2003) suggests that 1996 was an exceptional year, with much higher costs than
subsequent years. The average cost of collisions, including 1996, was estimated by
Cairney (2003) to be $21 million at all crossings, comprising $13.8 million at active
crossings and $7.1 million at passive crossings.
Table 6: Estimated collisions with different types of protection Type of protection 1996 1997 1998 1999 2000
Passive: Give-way 14.4 4.4 5.9 3.3 1.1 Passive: Stop - - - 1.0 5.2 Sub-total at Passive Crossings 14.4 4.4 5.9 4.3 6.3 Active 19.0 11.5 17.5 8.6 12.5 Other control 0.3 - - - - Total 33.9 16.0 23.5 12.9 18.7
N.B. Totals slightly exceed column entries due to rounding errors and small sums in some individual cells. This table excludes NSW.
2.4.5 International statistical comparison
Level crossings are a characteristic of railways in most countries of the world.
Developing countries are known to have very poor level crossing safety performance,
with Vietnam, Thailand and Bangladesh having the highest collision and casualty rates
in the Economic and Social Commission for Asia and the Pacific (ESCAP) region. The
countries of this region include but are not limited to India, Vietnam, Thailand, the
Philippines, Bangladesh, the Islamic Republic of Iran and the Russian Federation. A
report (United Nations, 2000) evaluating the cost-effective systems for level crossing
protection in this region found that:
Railways are ill-equipped to be in a position to monitor level crossing safety
effectively and to take both corrective and pro-active measures to improve the
safety of their level crossing installations.
(p1)
Few of these railway systems appear to have an adequate safety information
system, which would support any rigorous assessment of safety hazards and risks.
(p122)
35
In terms of developed countries, international comparisons indicate that Australia
performs rather well. According to an Austroads project requested by the Australian
Transport Council (2002b) that was timed to coincide with the ‘7th International
Symposium on Railroad-Highway Grade Crossing Research and Safety’, held at Monash
University in Melbourne in 2002:
Australia appears to have fewer fatalities per 100,000 population at level
crossings than New Zealand, the United States and Finland, and to have
considerably fewer fatal crashes at passively controlled crossings than New
Zealand and Finland.
(Cairney et al., 2002b)
This report also indicates that in the United States in 2000, there were 425
fatalities at level crossings (Hall, 2002), a rate of approximately 0.15 fatalities per
100,000 people. Finland experienced an average of 7.9 fatal collisions per year and 10.6
fatalities per year at passive crossings, which is equivalent to 0.15 fatal collisions per
100,000 population and 0.21 fatalities per 100,000 population. The majority of crashes
in Finland occur at passively protected crossings (74%) and therefore appear to have a
higher fatality rate than the United States (Pajunen, 2002, Cairney et al., 2002b). In New
Zealand, there are approximately 7 fatal collisions and 9 fatalities per year, with 0.18
fatal collisions per 100,000 population and 0.23 fatalities per 100,000 population
(Cairney et al., 2002b).
Schmid and Watson (2004) attempted to compare level crossing data from selected
countries around the world and examined different formulae that are commonly used to
evaluate the risk at level crossings. However, this analysis, presented at the ‘8th
International Level Crossing Safety and Trespass Prevention Symposium’ in Sheffield,
England, was fraught with errors. Unfortunately the corrected data remains unavailable.
36
Table 7: Comparison of fatal collision rates
* estimated on basis of proportion of collisions at passive crossings. Source: Cairney, P., Gunatillake, T. and Wigglesworth, E. (2002).
2.5 ENGINEERING MEASURES
2.5.1 Overview of engineering measures
With human factors being a relatively new area of research in transport safety,
particularly rail safety, many of the protection and warning devices used at level
crossings are based on tradition, rather than sound human factors research and design.
Despite protection and warning devices at level crossings being visually less effective
than protection and warning devices used at highway intersections (Green, 2002), rail
authorities do not have the required resources to upgrade such systems to meet
worldwide standards such as the American Railroads Standards (Afxentis, 1994).
However, level crossings, particularly passive crossings, present inherent dangers to
road users and as long as they exist, rail authorities will remain at risk of being found
liable, in either whole or in part, for collisions at level crossings (Stephen, 2002).
Engineering design and warning systems are the most common element for which rail
authorities are found to be liable (Stephen, 2002).
Cairney (2003) states that “the form of traffic control implemented at a railway
level crossing greatly affects the decision that has to be made by the driver of the road
vehicle and the safety of the crossing” (p1). Where devices have been upgraded,
engineering solutions have demonstrated some ability to optimise the intrinsic safety of
roads at level crossings. There are a variety of engineering factors that need to be
considered in order to reduce collision risk at level crossings. These include signage and
markings; rumble strips; sight distances; information handling zones; crossing surface;
37
approach speed limits; boom barriers; conspicuity of trains; train horns; low-cost
warning systems; vehicle dimensions and performance; and crossing closure. Each of
these factors is considered by risk based scoring systems that assess risk at individual
crossings.
Despite such warning devices alerting motorists to be drive safely at level
crossings, the behaviour of the motorist approaching a crossing can have a significant
effect on the probability that a collision with a train will occur (Carroll et al., 1995). A
variety of motorist errors may contribute to a collision between a motor vehicle and
train, despite the use of warning and protection systems at level crossings. These include
(but are not limited to):
• Failure to detect the train before it reaches the crossing;
• Failure to recognise the potential hazard of a train; and
• Failure to correctly estimate when the train will arrive at the crossing.
(Carroll et al., 1995)
In part, these errors can be traced to the quality of the train-related information
(visual and audible) needed by motorists to take appropriate action when approaching
and crossing a level crossing. Carroll et al. (1995) suggest that the visibility of the train,
configuration and geometry of the crossing, and the peripheral vision of the motorist, are
all factors that impact on a motorist’s ability to acquire the information necessary in time
to avoid a collision. Train visibility at passive crossings with no active warning devices
or protection systems, is a particular concern. As such, it can be said that no matter how
skilled or experienced a motorist is, the physical environment and engineering systems
of a level crossing may contribute to a collision between a motor vehicle and a train.
Therefore, it is essential that motorists are able to detect trains approaching or at a
crossing and recognise the associated dangers as early as possible in their approach to a
crossing, in order to avoid a collision.
As part of a cooperative approach in reforming the rail industry in Australia, the
Australian Rail Operations Units was established to work with industry and the States
and Territories to develop a national voluntary ‘Code of Practice for the Defined
Interstate Rail Network’, that aims to assist in harmonising operational and engineering
38
procedures across the industry. This code is currently being implemented in the rail
industry throughout Australia.
2.5.2 Signage and markings
The Australian Standard (1742.7-2007) for signing and marking treatments at both
passive and active crossings is very comprehensive. This Standard:
…specifies traffic control devices to be used to control and warn traffic at and in
advance of railway crossings at grade. It specifies the way in which these devices
are used to achieve the level of traffic control required for the safety of rail traffic
and road users, including pedestrians. Requirements and guidance are also given
in appendices on the illumination and reflectorisation of signs, on their
installation and location, and on selection of the appropriate sign size.
This current Standard draws attention to the need for meaningful cooperation on
both safety and maintenance issues at level crossings between relevant road and rail
authorities (Standards Australia, 2007). It clearly asserts that the safety of road users is
dependent on their ability to detect the approach of a train and that any changes in
infrastructure or operation contemplated by one authority that may increase the risk
associated with operation of the crossings, needs to be agreed by the other authority
(Standards Australia, 2007).
2.5.3 Rumble strips
Rumble strips are either grooves or rows of raised pavement markers that are
placed perpendicular to the direction of travel to alert drivers to an approaching change
of roadway condition or hazard requiring substantial speed reduction (Washington State
Department of Transportation, 2004). The intention of rumble strips is to provide both
an auditory and physical (vibration) stimuli that can be felt by both the driver and
vehicle occupants. The corrugation height for passenger vehicles can be quite shallow
(around 8mm), yet still produce significant stimulus in a vehicle (Rechnitzer, 2002). The
39
use of tactile surface texture on roadways is a common design measure, and varies from
'rumble strips' on the edge of roads to 'corrugated surfaces' on the road way. Figure 4
illustrates some examples of rumble strips (Harwood, 1993).
Figure 4: Example of types of rumble strips
The usage of rumble strips at level crossings in Australia have been used from
time to time, but this use has not been systematic nor do the installations appear to have
been evaluated (Cairney, 2003). There has been a proposal to install a low cost rumble
strip system at passive crossings in Australia, which is only active when a train is
approaching (Rechnitzer, 2002). This rumble strip system is intended to be used in
conjunction with existing visual warning signs on approach and at the level crossing.
This system has been designed due to the infrequency of trains and the absence of
protection systems at passive level crossings, to warn motorists when a train is
40
approaching or not (Rechnitzer, 2002). Activation of this system would occur by an
electronic signal transmitted from the approaching train.
A review of studies of rumble strips in reducing road crashes at road intersections
performed by Harwood (1993), showed that although studies generally show a reduction
in collisions, it is not possible to quantify their effectiveness due to small sample sizes or
inadequate methods. Additionally, Harwood (1993) reports that rumble strips have not
been as effective as speed reduction measures. In the Netherlands, the effectiveness of
rumble strips in reducing speed, were examined as part of a large-scale project to
improve safety at level crossings commissioned by Railinfrabeheer B.V. Observations
were conducted on video recordings at 16 level crossings in the before-and-after study.
These video recordings were analysed quantitatively by determining the speed of free-
driving passenger cars and by conducting a time analysis of non-stopping/stopping
behaviour of pedestrians, bicyclists and motorists (van der Horst, 2002). This study
concluded that transversal rumble strips do not appear to be effective in reducing the
speed of free-driving passenger cars, neither on 50 nor on 80km/h roads (van der Horst,
2002). Cumulative speed-distributions did not differ significantly before and after
installation of these transversal rumble strips (van der Horst, 2002).
2.5.4 Sight distances
The sight of an approaching train may be obscured from a motorist’s view for a
range of reasons. According to Ward and Wilde (1995) “limited sight distance along the
tracks will compound the motorist’s difficulty in detecting an approaching train and the
estimation of its rate of approach” (p33). Objects in the driving scene, signs other than
the crossing signs, vegetation, buildings/structures, are all possible visual distractions
that may take a driver’s attention away from an approaching train (Caird, 2002). Failure
to detect a train by a motorist is largely the result of sight distance, particularly at
passive crossings (Berg, 1982). Some motorists may be aware that they have a restricted
sight distance, however may continue to appraise the prevailing hazard but still engage
in compensatory modification of their approach behaviour. Approach behaviour is
therefore modified to sustain a degree of perceived situation risk (Ward and Wilde,
41
1996). On the contrary, many motorists do not recognise that sight limitations present a
problem when driving at level crossings. Although sight distance is regarded by most
experts as being an important engineering factor in level crossing collisions, Ward and
Wilde (1996) argue that “evidence that restricted lateral visibility at railway crossings is
hazardous has not been forthcoming in spite of its apparent plausibility”. There has been
numerous research that has found that sight distance and level crossing collisions are not
correlated (Russell, 1974b, Zalinger et al., 1977, van Belle et al., 1975). Ward and
Wilde (1996) suggest that “the lack of evidence to link restricted lateral visibility with
higher accident rates seems paradoxical in light of the intuitive relationship between
these two variables” (p63). Schoppert and Hoyt (1968) state, whilst commenting on this
paradox:
This does not seem logical: sight distance should be one of the most important
variables. If the driver cannot see the crossing and down the track an adequate
distance, then he and his vehicle are being expected to perform beyond their
physical limitations.
(p29)
According to Dewer (2002) adequate sight distance depends on the speeds of both
train and motor vehicle. Mortimer (1988) argues that train visibility is sharply reduced
at level crossings that are at an angle to the road, with a significant number of level
crossings in the United States having angles between 60 and 90 degrees (approximately
80%) and 4% having very sharp angles of less than 30 degrees. It has been demonstrated
that the risk of a collision is seven times higher when the road is parallel to the tracks
and the train is approaching from behind the driver’s field of vision (Caird, 2002).
Alexander (1989) proposes that the limits of head movement for scanning at level
crossings are considered to be 110 degrees to the left and 140 degrees to the right.
Therefore, truck drivers may be at a disadvantage, with the design of truck cabins
limiting views for drivers. Dewer (2002) suggests that trucks typically have more
difficulty stopping and accelerating, hence the need for greater sight distance to stop and
to cross the tracks ahead of the train after stopping at the crossing. Dewer (2002)
recommends that a minimum of eleven (11) seconds of travel time by the train should be
42
available to the truck driver in order to accommodate acceleration capability and length
of large trucks.
2.5.5 Information handling zones
According to Tustin (1986) the situation faced by a motorist of any vehicle at a
crossing occurs in three areas or zones. Information handling zones are particular areas
of the road that motorists make decisions about the level crossing ahead (Tustin, 1986).
The three zones include:
• Approach zone – This zone is the area of the road in which motorists begin to
formulate actions needed to avoid colliding with trains. Scanning for trains or
signals, recognising any hazards, and deciding the proper course of action, are
behaviours that motorists use in this zone. The motorist must be aware of the
crossing ahead, with information usually provided to the driver by an advance
warning sign or pavement markings. The driver must take notice of the crossing
through visual observations, control devices or sounds from the train horn
(klaxon). Advance warning systems should be place in an area that provides
sufficient warning to motorists to alter their speed and take appropriate driving
action as required.
• Non-recovery zone – This zone begins at the point along the road where
motorists must make a decision to stop (after stopping at a ‘Stop’ sign or giving
way at a ‘Give Way’ sign) if a train is approaching. If the stop/go decision is
delayed beyond the beginning of the non-recovery zone, the amount of road
remaining will be insufficient to avoid a collision with an oncoming train. The
non-recovery zone ends at the beginning of the hazard zone, and starts at the
stopping sight distance required by the vehicle speed. Proper design and
installation of warning systems and control devices will provide the majority of
drivers with the information needed to make the decision (in time) to stop if
required. Provided with such information, the motorist must operate their
vehicle as required by the prevalent conditions (e.g. visibility of an approaching
train).
43
• Hazard zone – This zone is the rectangle formed by the width of the road and
distance measured along the road on either side of the tracks. This zone is the
area where stopped or approached vehicles are capable of colliding with
approaching or stationary trains. The objective of this zone is for the motorist to
cross the tracks safely. Obeying warning signals and protection systems is
crucial for crossing safely in this zone.
(p31-33)
2.5.6 Crossing surface
Improving the quality of the surface on approach to level crossings is one such
treatment that initiates great debate. Vehicles that are aware of a crossing ahead, but are
unable to stop in time due to poor quality road surfacing, may be at risk of a vehicle-
train collision. According to Glennon and Engr (2005), rough crossing surfaces are
sometimes alleged to be a contributing factor in distracting motorists from their primary
task of searching for a train. It is unknown however the extent to which crossing surface
is responsible for collisions at level crossings. More research is needed into the
correlation between crossing surface and distraction at level crossings.
2.5.7 Approach speed limits
A field study conducted by Ward and Wilde (1995) was undertaken in Ontario,
Canada to compare motorist approach behaviour between day and night-time conditions.
All observations of approach behaviours were conducted without the crossing signals
being activated. Day and night observation periods were made within the same level
crossing and in close temporal proximity to minimise confounding and any migration of
motorist characteristics (Ward and Wilde, 1995). This study found that approach
behaviour to the crossing examined was far more conservative at night than during the
day. Motorists in this study had consistently lower approach speeds throughout the
approach way and navigated the crossing at a slower overall speed at night (Ward and
Wilde, 1995). Additionally, there was some evidence of less braking overall at night,
with Ward and Wilde suggesting that the slower speeds were indicative of a slower
initial speed prior to the approach way, rather than by subsequent modification of speed
44
through braking during the final approach. Ward and Wilde (1995) argue that these
differences could not be attributed to systematic demographic changes in day and night-
time motorist populations.
2.5.8 Boom barriers
A program was conducted in Victoria during a 19-year period (1971-1989) in
which 91 metropolitan and major urban level crossings were upgraded from flashing
lights only status to boom barrier status (Wigglesworth and Uber, 1991). There was
significant reduction in fatal vehicle-train collisions during this period, with the
mortality rate being reduced from 5.71 to 0.33 (deaths per 100 crossing years). This
study compared these reductions with control sites (consisting wherever possible of the
next crossing on the same line as the upgraded crossing), which saw mortality rates
increase from 1.22 to 1.63 deaths per 100 crossing years (Wigglesworth and Uber,
1991). A second comparison was also made in this study – between the same 91
crossings and the 82 crossings on the same lines where the flashing light installations
had remained unchanged for the complete 19-year period. The mortality rate for this
second comparison group rose from 1.31 to 2.77 deaths per 100 crossing years.
Although this study was conducted a few decades ago, it is clear evidence that upgrading
level crossings with flashing lights to boom barrier status reduces fatal collisions at level
crossings (Wigglesworth and Uber, 1991).
2.5.9 Conspicuity of trains
Alerting devices that enhance train conspicuity, allowing attraction of motorist
attention, increase the likelihood that a motorist will see the approach of a train in
sufficient time to safely cross at a level crossing. There is no doubt that under certain
conditions, the failure of a motorist to detect an approaching train is a major contributing
factor in vehicle-train collisions. Passive crossings without flashing lights or boom
barriers make it difficult for motorists to detect a moving train and correctly estimate its
time of arrival to the crossing. With the majority of level crossings in Australia being
passively protected, it can be difficult for motorists to detect a train at night. One
45
important factor in the failure of motorists to detect an approaching train is the lack of
visual properties on the train, other than its standard headlight (Carroll et al., 1995).
Although the majority of vehicle-train collisions in Australia occur during daylight
hours, when normalised for differences in traffic volume between day and night periods,
collision rates are most likely to be substantially higher at night than during the day
(Darzentas and McDowell, 1981, Leibowitz, 1985, Russell, 1974b, Schoppett and Hoyt,
1968, Swedish Road and Traffic Research Institute, 1981). According to Ward and
Wilde (1995), it is apparent from differences between daytime and night-time collision
rates that there are qualitative differences between (1) the conditions encountered by
motorists during these different periods at level crossings and (2) the style of approach
behaviour enacted by motorists (p32).
Conspicuity depends primarily on the contrast between an object and its
background to make an object obvious. At night, lighting on trains assists with
achieving such a contrast. Historically, standard train headlights have provided a visual
signal to motorists of an approaching train at level crossings, however headlights were
not specifically designed for that purpose (Carroll et al., 1995). Strobe lights, ditch
lights, crossing lights, oscillating devices, paint schemes and reflective materials, have
all been equipped on locomotives to increase train conspicuity (Carroll et al., 1995).
Train conspicuity research has been conducted since the 1970’s, and since this time
there has been many technological changes, both in the vehicle’s internal environment as
well as in alerting device material technology and techniques (Carroll et al., 1995).
During the mid 1990’s, the Volpe National Transportation Systems Center (Volpe
Center) evaluated the effectiveness of available auxiliary external alerting devices that
may improve train conspicuity at level crossings (Carroll et al., 1995). This study
evaluated a variety of external visual alerting devices including several light systems, as
well as paint schemes and reflective materials. The Volpe Center used a multifaceted
methodology to evaluate how external alerting devices contribute to the ability of a
motorist to:
• Detect the approach of a locomotive before the train reaches the level crossing;
• Recognise the associated potential hazard; and
• Accurately estimate when the train will arrive at the level crossing.
46
Results from this evaluation indicated that alerting light design and operation can
improve train conspicuity. Other results were found:
• Passive alerting devices are considered to be of only limited effectiveness in
enhancing train conspicuity. Passive alerting devices should be used only as a
secondary technique to reduce collisions at level crossings.
• The use of auxiliary external alerting lights can be effective in improving train
conspicuity.
• Multiple lights, light intensity, spatial dimensions and angle, and pattern all
contribute to increasing the effectiveness of the visual alerting signal and
providing more attention by motorists.
• Train approach speed, sight distances and ambient light conditions should be
considered when specifying minimum and maximum levels for alerting light
luminous intensity and effective intensity.
• The provision of a low-beam intensity control which supplies a lower luminous
intensity level for the entire alerting light system, similar to a ‘dimmer’ switch
currently used for standard headlights, would reduce the potential for glare.
A ‘cross-eyed’ alerting light beam pattern with lights angled inward
and focused an extended distance down the track, appears to have the
positive features of a wider system beam width and range in front of
the train as well as less potential for blinding motorists.
(Carroll et al., 1995)
In Australia, controlled field trials of auxiliary alerting lights have been conducted
to evaluate the most effective warning to motorists (Hughes and Coles, 1986). These
trials concluded that a 100 watt combination driving/fog lights whereby two lights are
angled 7.5 degrees inward (cross-eyed) provided ample warning to motorists and
improved track illumination directly ahead and to the side for the train driver (Carroll et
al., 1995, Hughes and Coles, 1986). A Western Australian field study that reviewed
three lighting systems [(i) locomotive headlights alone, (ii) locomotive headlights and
ditch lights and (iii) locomotive headlights, ditch lights and a single roof-mounted strobe
light], concluded that neither of the two auxiliary lighting treatments improved the
47
conspicuity of the locomotive over that achieved by the standard headlights (Cairney et
al., 2002a). The results from this study indicated that a single strobe light did not
improve detection when added to trains already fitted with both headlights and crossings
lights (Cairney et al., 2002a). More recently, an Austroads report examined the prospects
for improving the conspicuity of trains at passive crossings (Cairney, 2003). This report
highlighted the importance of addressing the critical aspects of vehicle-train collisions at
passive crossings for train conspicuity, which included:
• Daytime collisions are approximately 70% of the problem, night-time collisions
approximately 30%; and
• A vehicle is struck by a train in 65% of cases, and a vehicle collides with the side
of the train in approximately 35% of cases. In fatal collisions, a higher
percentage (approximately 80%) involved the vehicle being struck by the train.
(Cairney, 2003, p29)
Cairney (2003) suggests two possibilities for addressing the issue of daytime
collisions. Additional lighting is one such possibility, however there is limited research
that has tested the effects of coloured strobe lights on the daytime conspicuity of trains,
and Cairney (2003) questions whether there are any suitable colours that have not been
preempted for other uses. The second issue is that of the effects of colour schemes
which contrast with the backgrounds against which a train is seen. Again Cairney (2003)
suggests that the effect of paint schemes in enhancing the brightness contrast they can
offer with the environment is limited and no data relating to train conspicuity is
available. With regards to collision reductions, this report suggests that “evaluating the
effectiveness of conspicuity treatments to trains in terms of collision reductions will not
be practical, due to the small number of collisions available for comparison, unless the
proportion of collisions prevented by the treatment is exceptionally high” (Cairney,
2003, pix). However, due to there being a smaller number of trains (approximately
2300) than passive crossings (approximately 6000), alerting devices on trains are likely
to cost less than installing low-cost active warning systems at passive crossings
(Cairney, 2003).
48
The 2002-2003 Annual Report of the Department of Transport and Regional
Services refers to the implementation of a number of key strategies to support safety
transport services; lead the development nationally of more consistency in rail regulatory
arrangements; and develop the Commonwealth’s role in rail safety and investigation.
Arising from this report, the House of Representatives (Standing Committee on
Transport and Regional Services) inquired into some of the measures that have been
proposed to improve train conspicuity and reduce level crossing collisions. From this
inquiry in June 2004, the House of Representatives recommended that:
The Australian Government takes steps, through the Transport Ministers Council,
to require that all locomotives and rolling stock in the Australian rail industry are fitted
with standard reflective strips or reflective paint and that all locomotives are fitted with
rotating beacon lights.
(2004, p12)
The Australian ‘Code of Practice for the Defined Interstate Rail Network’
specifies for lighting that:
• Headlights
Headlights may be fitted in a single central assembly, or as two
individual headlights positioned each side of the centre line of the
locomotive.
Each headlight should be protected by a separate circuit breaker, and
controlled by switches for off/on, main and dim positions.
Each lamp should be supplied from a separate circuit. The two units
should not use common components except for the headlight circuit
breaker and control switches.
Each headlight should incorporate at least two lamps, and should have
provision for adjustment in both the vertical and horizontal planes.
• Coupler lights
Protected coupler lights are optional equipment but should be
provided to illuminate the front of the locomotive in the vicinity of the
49
coupler and the ground below. They should not project light
forwards.
Switches for these lights should be located on both sides of the
locomotive at both ends and should be accessible from track
formation level.
• Road visibility lights
Low level lights should be provided on the ends of each locomotive to
provide road users and pedestrians with enhanced warning of the
presence or approach of a locomotive or train.
These lights should be positioned below the anti-climb beams and
near the outside of the underframe, and should be capable of
adjustment in both the horizontal and vertical planes. Each light
should normally be set to face in towards the track centre line; the
included angle between the lights shall be 10 – 15 degrees.
They should be protected from incidental damage under normal
operating conditions.
The lights should not be less than 100 Watts each and may be
combined with or incorporated into the fog or low visibility lights.
(Commonwealth of Australia, 2002, p50-51)
Illumination at crossings is another treatment option for improving conspicuity of
trains at level crossings. Detection of a slow moving train is far more difficult than a
fast moving train. This is one reason that illumination at crossings is paramount. Some
research in the U.S.A. has shown that crossing lights were the most effective treatment
of all lighting treatments for detection of trains at a level crossing (Federal Railroad
Administration, 2001). Other research has found that benefits of night-time illumination
at level crossings is approximately 30%, with benefits being somewhat greater at passive
crossings (Russell and Konz, 1980). It has also been suggested that greater illumination
is required at crossings that have greater frequency of trains which are slow or stopped at
crossings at night (Dewer, 2002). As train speed increases, the chance of a vehicle-train
collision decreases. Although the majority of vehicle-train collisions in Australia occur
50
during daylight hours, when normalised for differences in traffic volume between day
and night periods, collision rates are most likely to be substantially higher at night than
during the day (Darzentas and McDowell, 1981, Leibowitz, 1985, Russell, 1974b,
Schoppett and Hoyt, 1968, Swedish Road and Traffic Research Institute, 1981).
2.5.10 Train horns
Train horns (klaxons) are an important part of a train’s safety procedures, as it
serves as both an early warning for motorists as well as emergency alarm to convey
urgency to motivate a motorist to correct their action and clear the crossing (Russo,
2003). However, there has been little research into the effectiveness of such warning
devices for motorists, particularly with regards to the ambient noise inside the vehicle
being dominated by noise resulting from the vehicle’s operation (Rapoza, 2002). Much
of the literature has focused on noise levels to address complaints by residents who live
near tracks. A study by the Rail Safety Directorate at Transport Canada investigated
horn placement on locomotives and emitted sound to address excessive loudness
complaints from crews and residents near tracks (Transportation Development Centre,
2003). This study found that the location of a train horn was very important to its
effectiveness and a horn’s harmonic content is more important than its fundamental
frequency (Transportation Development Centre, 2003). An investigation for the Rail
Safety and Standards Board in the U.K. found that the environmental impact of noise
from the warning horns could be reduced with the horns still clearly audible at 400
metres by using devices that are 6 dB quieter than those currently specified for speeds of
160 km/h or less (Hardy, 2004). Another study investigating how to increase the
effectiveness of train horns without increasing intensity found that “train horns could be
made more effective by ensuring substantial mid-frequency energy, shifting the spectral
centroid higher, and increasing musical dissonance” (Russo, 2003, p51). In terms of the
effect that bans on train horns have on level crossing collisions, Zador’s (2003) study
found statistically significant and substantial increases in collisions (ranging from an
increase of 43.3% for active crossings with flashing lights to 52.6% for passive
crossings). However, other studies do not support these findings.
51
According to a study conducted by Aurelius and Korobow (1971) a train horn
would be an unacceptable nuisance if it had enough output to be totally effective in
warning motorists. Their study found that:
• The sound pressure level to alert a motorist (reliably) traveling at up to 50 miles
per hour (80 kilometres per hour) was 105 decibels;
• The warning level immediately outside the vehicle rose to 109 decibels, at speeds
between 50 and 65 miles per hour (80 to 105 kilometres per hour);
• High figures make no provision for age-related, occupation-related, or leisure-
related activities resulting in hearing loss, nor do they allow for the effect of
gusting wind; and
• Present horns cannot reliably warn motorists early enough of an oncoming train
(in high speed encounters).
Wigglesworth (2002) supports this view, proposing that enhancing the train horn is
an unacceptable nuisance and there is little point in further research “unless new
technology intervenes in some way” (p5).
2.5.11 Low-cost warning systems
Many motorists are “accustomed to driving in a well-engineered and carefully
controlled traffic environment with prior notification of any threat” (Wigglesworth,
2002, p9). When motorists (who live in urban areas) travel to more rural areas, they
expect to be told if a train in approaching a level crossing. There are studies that support
this argument. A study by Sanders (1976) found that 35% of Californian drivers thought
all level crossings had active warning systems. Installing flashing lights or boom
barriers (active protection) at all passive crossings in Australia, is not cost-effective for
the number of fatalities per year. Wigglesworth estimates that to install active protection
(approximately $300 000 per crossing) at all public passive crossings in Australia
(approximately 6000), this would cost approximately $1.8 billion. Government
authorities in Australia are unable to provide such funding to upgrade all level crossings
52
to active protection. Therefore, one alternative is to develop and evaluate installation of
low-cost warning devices at passive crossings. According to Jordon (2006)
With the high costs associated with traditional active level crossing protection,
many traffic engineers in the 1970’s and 1980’s expressed frustration at the
inability to be able to fund active protection at many of the passive crossings that
were the scenes of fatal crashes.
(p1)
The rationale for low-cost warning devices is basically to stretch the dollar further.
With many of these passive crossings being in isolated rural areas (with small road
traffic and train volumes), rail industry budgets typically are not able to provide more
active protection systems. In the early 1990’s in Victoria, a group of engineers united
to originate the Victorian Rail Level Crossing Committee. This committee’s interest in
alternative level crossing protection systems, alongside several Coronial Inquests into
fatalities of motorists at passive crossings in Victoria’s north, added weight to the
concerns for safety at passive crossings (Jordon, 2006). According to Jordon (2006) the
intention of such devices is “ not to replace the existing active device with a new low
cost option, but rather to have an additional device which could be used to improve the
conspicuity of selected passive crossings at the time that a train was in the vicinity” (p1).
The major reason for the high cost of warning systems is that track circuits are
electrically isolated sections of rail which are placed in electrical contact with other
sections by the passage of a train (Cairney et al., 2002b). “Modern sensing and
communications technology offers the possibility of lower-cost options for active
signaling which do not depend on track circuits for train detection” (Cairney et al.,
2002b, p16).
In 1999, a trial of five low-cost warning devices took place in Melbourne by
VicRoads and VicTrack, with the aim to improve road safety at remote rural rail
crossings by improving the conspicuity of the crossing at the time that a train is
approaching. The five devices included: Doppler radar unit, two magnetometers, an in-
train transmitter, and an induction loop. As a result of this trial, a Low Cost Level
Crossing Warning Device (LCLCWD) has been developed in conjunction with a
53
commercial partner that allows train detection to be achieved by loop detectors placed
between the tracks (Jordon, 2006). This device has robust detectors and radio linked
information transmission, is ideally suited to rural locations remote from mains power,
and is environmentally sustainable technology (Jordon, 2006). Additionally, it is
powered by solar cells with battery storage, which allows substantial savings (Cairney et
al., 2002b). In late 2004, the first real-life installation of the device took place. This
device has been undergoing ‘blind’ testing under full train conditions near Creswick for
nearly a year (Jordon, 2006). VicRoads and VicTrack are planning for this device to
soon become an accepted treatment for passive crossings in remote areas throughout
Victoria.
2.5.12 Vehicle dimensions and performance
The variety of vehicles that cross level crossings must be taken into account in the
design and operation of crossings (Tustin, 1986). Motorcycles, tractors, bicycles, trucks,
buses, vans, are types of vehicles other than cars that may travel over a level crossing.
These vehicles have widely different characteristics that may directly affect the design
elements of a crossing (Tustin, 1986). Of equal importance is the cargo/freight that
these vehicles carry (such as children in buses or hazardous materials in trucks) (Tustin,
1986). According to Tustin (1986):
Long vehicles, and vehicles carrying heavy loads, have longer braking distances
and slower acceleration capabilities; hence, long vehicles may be exposed to a
crossing for an even greater period of time than that in proportion to their length.
(p35)
Tustin et al. (1986) also suggests that the width of the vehicle may be a factor in
considering the width of the road surface and the length of the crossing surface
measured along the track. Trucks with greater widths are now becoming more
commonplace on Australian roads. The combination of under-clearance and wheelbase,
is another important component that is required to be considered in the design of level
crossings. Long trucks with low clearance may become lodged on a crossing if the
54
grades of the crossing and its approaches are not adequate (Tustin, 1986). Braking and
acceleration performance of a vehicle is another issue pertinent to level crossing safety.
The function of a vehicle’s braking performance is subject to stopping sight distance.
If a crossing experiences a significant percentage of heavy trucks, any given sight
distance will dictate a slower speed of operation to allow for the braking
performance of these vehicles.
(Tustin, 1986, p35)
The acceleration of vehicles is also of key importance. Clearing the crossing
before a train reaches the crossing, is of particular importance for trucks. Coupled with
their long lengths, trucks have relatively poor acceleration capabilities, which make
them hazardous at passively protected level crossings. The design and condition of the
crossing surface, is also considered an important safety aspect (Tustin, 1986). Crossings
and approaches built on a steep rise are often difficult and time consuming for vehicles,
let alone trucks, to cross (Tustin, 1986). All of these factors must be taken into account
by rail authorities when B-double approved routes are applied for by the heavy vehicle
industry.
2.5.13 Crossing closure
From the literature reviewed, it is evident that level crossings present a major
hazard to motorists and are typically the greatest cause of fatalities and injuries in rail
operations. The first treatment option at hazardous level crossings should always be
closure of the crossing. Crossing closure can be achieved either by: grade separation
(bridge or tunnel), closing the crossing to road traffic, or closing the crossing to rail
traffic through relocation or abandonment of the rail line (Glennon and Engr, 2005). In
considering the need and prioritising crossing closure, screening tools are required. In
Australia, closure of a level crossing is assessed by use of the Risk Based Scoring
System. Rail authorities actively seek to close level crossings, both public and private,
wherever possible. Level crossings can be nominated for closure by the score, local
governments, Main Roads department, the rail industry or the general public.
55
2.5.14 Risk based scoring systems
Models to assess risk at individual level crossings have received great attention
during the past few decades, with models existing for more than 60 years (Mok and
Savage, 2005). A comprehensive system for assessing risks at level crossings in
Australia and for assessing the likely impact of different treatments on these risks was
developed by Queensland Rail in 1999. The Queensland Level Crossing Safety Project
Team researched available risk scoring systems for level crossings, and found that there
was no single risk scoring system that was satisfactory for the purpose of assessing
safety risk at level crossings within Queensland. Therefore, they developed a new
system that used the best features of the systems that the team reviewed. This resulted in
the Australian Level Crossing Assessment Model (ALCAM) (Hughes, 2002). The
ALCAM is a process that objectively assesses, evaluates and prioritises the safety risks
of level crossings. It is also a method to determine the best treatment for individual sites.
This model allows for considerable improvements in safety at level crossings,
particularly at passive crossings where the use of common active treatments is not
possible (Hughes, 2002). Risk assessment for any given crossing involves examining the
matrix formed by the risk factors (29 possibilities) and the possible collision mechanism
(19 possibilities) (Cairney, 2003). The risk assessment consists of a total score (the sum
of all the cell entries) plus a score for each of the risk factors and a score for each of the
possible collision mechanisms. After determining that a level crossing is high risk and
what mechanisms are contributing to the risk, the system then becomes a valuable tool in
assisting to determine the treatments that are appropriate to reduce the risk (Hughes,
2002). The model allows for risk reduction strategies to occur in one of the following
ways:
• By installing new treatments at the level crossing (e.g. improved signage);
• By changing the characteristics of the level crossing (e.g. changing the road
speed or improving sight distances); or
• By changing road and the rail vehicle volumes at the level crossing (e.g. by
diverting road traffic to a nearby bridge over the railway line).
(Hughes, 2002)
56
This model also has the capacity to assess benefit/cost for proposed improvement
works, ensuring that each dollar is spent where it can generate the greatest safety
improvement. The factors include:
• Road and rail volumes and speeds at existing level crossings;
• Safety treatments;
• Number of tracks;
• Road quality;
• Vision of road vehicles to train;
• Vision of rail vehicles to road; and
• Adjacent road geometry.
(Williams and Creber, 2005)
It is important to note that this scoring system does not use information regarding
previous collisions that have occurred at level crossings to calculate a risk score,
although it has been found that the system produces results that have a very strong
correlation with collision history (Hughes, 2002). In May 2003, all Transport Ministers
in Australia agreed to adopt this innovative method of risk assessment. Additionally, this
model has been adopted by other countries such as the United Kingdom.
2.6 LAWS AND ENFORCEMENT
2.6.1 Overview of enforcement measures
The basic objective of enforcement is to ensure that the perceived costs of
infringing a law outweigh the perceived benefits of doing so…Research on traffic
law enforcement tends to indicate that the absolute size of the sanction is less
important than the subjective probability of apprehension. In other words, the
threat of detection tends to be a more effective deterrent than the severity of the
punishment.
(Pickett, 1996, p40)
57
2.6.2 Level crossing laws in Australia
2.6.2.1 Australian road rules
The document ‘Australian Road Rules’ approved by the Australian Transport
Council and published by the National Road Transport Commission in 1999, provides
rules to be followed by all road users. They are part of a national scheme to provide
uniform road laws throughout Australia (National Road Transport Commission, 1999).
2.6.3 Enforcing laws at level crossings
2.6.3.1 Traditional methods of enforcement
Traditional methods of police enforcement for red light running involve police
patrols. However, police resources are typically limited in the amount of traffic
enforcement that can be conducted. According to Porter and Berry (2001) motorists
learn to slow down and obey the road rules when they see police, but they are aware that
enforcement is irregular and inconsistent. The traditional method for a driver violating
level crossings to be ticketed is if a nearby police officer observed them or a train driver
recorded their details and forwarded the information to police. The number of crossings
in Australia that have histories of violations and/or collisions makes the deployment of a
police officer at crossings both economically infeasible and practically impossible due to
limited police resources. As such, enforcement programs must be both cost-effective
and undisruptive to natural traffic flow.
Police presence may cause distractions at complex level crossing intersections
(Fitzpatrick et al., 1997), with automated enforcement technology being the most
practical solution at active level crossings. In smaller towns where more passive
crossings exist, police patrols at level crossings may have some deterrent effect. No
studies have been carried out in Australia that investigate any traditional enforcement
programs at level crossings. Although studies for other driving behaviours have found
positive results with enforcement programs, such results can not be generalised to other
driving behaviours as they do not produce similar results (Tay, 2005).
58
2.6.3.2 Automated enforcement technology
“Automated enforcement typically consists of detection equipment and cameras to
capture images of drivers violating the law” (Lammert, 1999, p284). This information
technology countermeasure works in a similar way to red light cameras as it takes a
photo of any vehicle that illegally crosses the level crossings. Some research has
indicated that such devices improve safety at level crossings. In a trial in California,
level crossing enforcement systems cut the violation rate by 78-92% (Rathbone, 1995).
Pickett and Grayson (1996) suggest that an integrated program of red light cameras
coupled with high profile police campaigns offer the greatest change of success in
reducing red light violations at level crossings.
As part of a study by the Department of Infrastructure in Victoria (Public
Transport Safety Victoria, 2005), a Redflex Traffic Camera System was installed at
Springvale Road, Nunawading to observe motorist behaviour at this busy crossings.
This crossing was chosen for numerous reasons:
• Train drivers regularly report near miss incidents at the crossing;
• High volumes of rail and road traffic;
• Adjacent to a major road intersection; and
• Frequent road congestion at the railway line.
(Public Transport Safety Victoria, 2005)
This study found that on average more than 200 incidents of motorist non-
compliance with the road rules occurred each weekday, 116 incidents occurred on
Saturdays and 101 incidents occurred on Sundays. The three main high risk behaviours
taken by motorists included:
• Entering the crossing while the warning lights were flashing, warning bells were
ringing or boom barriers were closing;
• Entering the crossing while warning lights were flashing, warning bells were still
ringing and boom barriers were opening; and
• Entering the crossing even though the crossing or the road ahead was blocked
and it was not possible to drive through and to clear the rail tracks without
stopping.
59
Of most concern with these findings is the frequency with which motorists stopped
on the crossing when it was blocked due to congestion ahead of the crossing. This
behaviour occurred at all times including when the flashing lights and audible signals
were indicating the approach of a train (Public Transport Safety Victoria, 2005). The
overall goal of this study was to provide footage to identify times and days when risk
taking behaviours by motorists are most likely to occur so that police are deployed to
deter traffic offences.
In New South Wales, RailCorp currently utilises ‘closed circuit television’
(CCTV) cameras across its network. These cameras record images of vehicles and
pedestrians using the crossing and signs at the level crossing notify users if a camera is
in operation. According to RailCorp, it collects, uses, stores and discloses CCTV images
for the purpose of:
• Monitoring and recording any offence or infringement, which may be reported to
the Roads & Traffic Authority or appropriate law enforcement agencies;
• Investigating safety/security incidents on or about level crossings;
• Monitoring the ongoing safety on or about level crossings and assess the need for
additional safety or maintenance measures; and
• Statistical or research purposes related to level crossing safety.
Images taken by RailCorp’s CCTV may be disclosed to the New South Wales’
Roads and Traffic Authority (RTA) or other law enforcement agencies such as the New
South Wales’ Police Service or Australian Federal Police (CityRail, 2005). However,
there have been no evaluation studies as to how cost-effective the installation of such
CCTV’s have been in reducing collisions at crossings in either New South Wales or any
other jurisdiction.
2.6.4 Fines and penalties
2.6.4.1 Overview
The fines for violating road rules at level crossings have historically been very low
in some jurisdictions in Australia in view of such a high risk interface between that of
60
rail and road. Previous to March 2007, the fine for disobeying road rules at level
crossings in Queensland was $45. Since 1st March 2007, Queensland Transport
introduced tougher penalties for risky driving at level crossings, with fines now being
$225 and 3 demerit points. Fines for disobeying road rules at level crossings are slightly
higher in New South Wales ($300 and 3 demerit points). Up until June 2007, Victoria’s
fines were lower than that of Queensland’s. However, as a consequence of the Kerang
train disaster in early June 2007 in which a semi-trailer collided with a V/line train
(killing 11 passengers and injuring at least 22), the Victorian government announced a
toughening on penalties for infringements at level crossings. These penalties have risen
from $177 and 3 demerit points to $430 and four demerit points. Additionally, a new
offence has been introduced for speeding to ‘beat a train’, crossing tracks when lights
and bells are operating, or weaving in between boom gates that have descended. Such
behaviours now carry a fine of 30 penalty units ($3304), 4 demerit points and automatic
3 month licence suspension.
2.7 EDUCATION, PUBLICITY AND TRAINING
2.7.1 Overview
It can be said that no matter how skilled or experienced a motorist is, the physical
environment and engineering systems of a level crossing may contribute to a collision
between a motor vehicle and a train. In light of the ongoing research and experience of
the systems approach to road safety, the role that road user education, driver training and
publicity campaigns play in influencing safe driving behaviour is being re-examined
(World Health Organization, 2004). Motorists have a tendency to choose a certain level
of performance that depends on how they wish to cope with the task demanded of them,
which makes the distinction between performance and behaviour important in the
examination of such road safety interventions (Henderson, 1991, Haid, 2002). The
World Health Organization suggests that when used in support of legislation and law
enforcement, education, publicity and training can create shared social norms for safety
(World Health Organization, 2004). As there is currently no effective model for the
61
relationships between how people drive and how they learn to do so safely, road safety
educational countermeasures must be developed through carefully designed human
factors and educational research (Henderson, 1991, Sentinella, 2004).
2.7.2 Driver training
Driver training is one area that has received great criticism. Driver training is
generally perceived by the public as being able to improve young driver’s behaviour.
According to Simpson (Simpson, 1995) there is a public perception that to attain its loss
reduction potential, the focus of driver training should be on those aspects of the driving
task that are linked to the risk of collision. However, research suggests that driver
training of a traditional and conventional nature contributes little to reductions in risk
among drivers of all age and experience groups (Mayhew, 1996, Woolley, 2000,
Christie, 2001). Christie (2001) argues that “the degree to which driver training can
address the negative influence of driver age and experience on crash risk and
involvement is probably limited…..and driver training for learner and novice drivers
may contribute to increased crash risk by encouraging early driver licensing” (p6).
Additionally, Haworth et al’s (2000) Australian research supports this view stating that
“despite the strong belief in the effectiveness of driver training courses by those
involved, there is no clear evidence that they are effective in lowering crash rates” (p18).
International studies reveal similar results. A well-known review of 30 scientific
evaluations of formal driver training, motorcycle rider training programs and advanced
training courses for novice drivers, found little support for the claim that driver training
is an effective countermeasure (Mayhew, 1996). Although it is important to examine
driver training as a road safety countermeasure, this form of countermeasure is not
within the scope of this thesis.
2.7.3 Publicity
The use of intensive road safety publicity campaigns are adopted by most
jurisdictions in Australia. Publicity, also known as public education, is usually
employed through television, radio, print media, press conferences and displays. Such
62
publicity campaigns usually have mass media advertising as the most visible component
of the campaign. Speeding and drink-driving are generally the subjects of such
campaigns, due to these behaviours being leading contributing factors in fatal road
crashes. The Global Road Safety Partnership, a hosted program of the International
Federation of Red Cross and Red Crescent Societies (IFRC) based in Geneva, suggests
that there are three main types of campaigns:
• To raise awareness of an issue or to inform (e.g. about new laws);
• To change attitudes (e.g. to improve public acceptance of road safety
countermeasures); and
• To change behaviour, as part of a package of measures (e.g. engineering and/or
enforcement related to speeding).
(Global Road Safety Partnership, 2006)
The effectiveness of such campaigns has been the topic of great debate for many
years. Typically, measures of whether such campaigns are effective have been confined
to recall of the message or assessment of how the public received the message
(Henderson, 1991). However, Avery (1973) proposed that neither reception nor recall
were related to whether or not safety-related behaviours had been modified by the
message. Christie (2002) argues that positive change in behaviours has been reported
when the information consists of only a simple, single message (e.g. use a child restraint
and/or install a smoke alarm).
The continuing debate in Australia has been fuelled by the Victorian Transport
Accident Commission’s (TAC) use of highly emotive and dramatic road safety
advertising. The TAC campaigns have been extensively evaluated with mixed results
(Cameron, 1993, Newstead et al., 1995, White, 2000, Cameron, 2000, Tay, 2004).
Some studies have observed no significant impact on driver safety as an indirect result
of publicity campaigns, whilst other researchers have found campaigns to be extremely
effective in reducing road trauma. Donovan (1999) evaluated 12 television commercials
(four drink driving, four speeding, two fatigue and two inattention) selected from a pool
of approximately 100 commercials from Australia and New Zealand (including TAC
ads). The study design used for this evaluation was a monadic, independent sample, with
63
potential target audience respondents being intercepted at suburban and city shopping
centres. The results indicated that there was no direct evidence of the inherent
superiority of dramatic, highly emotional ads (such as the TAC ads), over other less
dramatic, low emotion ads (Donovan, 1999). Additionally, the study found that several
of the 12 ads had substantially higher impact as a passenger than as a driver. Although
this evaluation used ‘likelihood’ dependent variables for analysis (i.e. (1) likelihood of
complying with the recommended behaviour as a driver, (2) and as a passenger, the
likelihood of influencing the driver to comply with the recommended behaviour),
Christie (2002) suggests that there are significant limitations to this evaluation by not
using injury and/or crash (or collision) reduction as outcome measures.
Common methodological limitations associated with evaluating the effectiveness
of campaigns have been highlighted by Christie (2002). Christie (2002) claims that the
effectiveness of publicity campaigns has generally received a poor record in crash (or
collision) and injury reduction terms. Evaluation of campaigns have rarely used injury
and/or crash (or collision) reduction as outcome measures, using acceptance, popularity,
and/or improvements in knowledge and skill as criterion measures (Duperrex, 2002,
Christie, 2002, O’Neill, 2002). Other common methodological limitations include the
simultaneous investigation of several variables using a correlational design, and the
reliance on limited outcome measures. Another explanation that may account for the
limited effectiveness of educational programs found in some research, relates to the
underlying effectiveness of the campaign development. Often campaigns are
commissioned by advertising agencies and are developed based on intuitive aspirations
rather than being guided by theoretical models and evaluative research.
According to a report to the TravelSafe Committee in Queensland, on achieving
high levels of compliance with road safety laws, Elliott noted that “missionary zeal
and/or a desire to ‘alleviate the carnage’ often means pressure applied to road safety
authorities and to politicians to do something” (Elliott, 1992). Elliott’s advice to this
TravelSafe Committee also included resisting temptation to just “do something” and
continue to apply proven methods of enforcement supported by publicity (Elliott, 1992).
The combination of supporting publicity with enforcement is not disputed by
experts. Homel et al (1988) concluded in their study of specific and general deterrence,
64
that both mass media and visible enforcement are important for campaigns to have any
significant impact. However, there are areas in road safety that make the justification of
the use of police resources difficult. Level crossing safety is one such area. Therefore,
educational interventions designed using theoretical models and evaluative research,
allowing the targeting of existing perceptions, attitudes, beliefs and behaviours, is seen
to be a more cost-effective method of attempting to change intention of motorists at
level crossings.
2.7.4 Education programs and interventions
Education of the general public (publicity) and education of specific road users are
two distinct forms of road safety education, although they are both often referred to as
‘road safety education’ in the literature. Targeted education interventions/programs, on
the other hand, are principally designed with specific road user groups in mind. These
interventions/ programs are usually designed using best-practice principles of road
safety education (Elkington, 2003). With motorist error being considered an important
factor in the causation of road crashes, great emphasis has historically been placed on
reducing error through education programs. However, with the recent emphasis on a
systems approach to improving road safety, the reliance solely on education has been
questioned. This debate can be resolved by examining the effectiveness of such
programs by considering the empirical evidence. However, an exhaustive review of the
literature established that there is very little research that examines the effectiveness of
educational interventions for specific road user groups.
Di Pietro and Ivett (2003) propose that “the contribution of road safety education
and training as a road safety countermeasure should not only be evaluated against
immediate outcomes relevant to the public health agenda, but also against the future
requirements of maintaining positive health outcomes” (p1). They suggest that such
countermeasures play an imperative role in preparing road users by developing skills
and understandings, which contribute to the development of a ‘road safety culture’ in the
community. Their argument includes the notion that “whilst it is sensible to fund
countermeasures which will rapidly reduce the road toll, it is equally important that
65
adequate funding also be allocated to education and training as part of a longer term
strategy” (p1).
Pedestrian education is one road safety area that has received positive results in
terms of education. A recent and extensive review of pedestrian safety in developed and
developing countries showed that it can change observed road crossing behaviour,
although its ability to reduce the risk of pedestrian injury in road crashes is largely
unknown (Duperrex, 2002). Another study (Leadbeatter, 1997) that found positive
results is the evaluation of the ‘RoadSmart’ program. RoadSmart is a primary school
road safety education program developed by VicRoads. Leadbeatter (1997) evaluated
the ‘Walking Safely’ component of the program and found that if teachers implement
the program in the intended manner, children cross roads more safely. A study
examining a program of educational and environmental (access prevention)
interventions designed to reduce the incidence of illegal and unsafe crossings by
pedestrians of the rail corridor at a suburban station in New Zealand (Lobb, 2001), found
that awareness-inducing interventions appear to increase knowledge of safe behaviours,
yet it is unclear as to whether they reduce injury rates (Lobb, 2001).
The argument against developing and implementing education
interventions/programs is largely led in Australia by Dr. Ron Christie a well-known
psychologist and road safety consultant. Christie (2002) argues that there is little
scientific evidence that road safety education contributes to reduced risk, injury or
fatality among those targeted. He suggests that such measures divert funds, resources
and attention away from better-based and more effective countermeasures (Christie,
2002). He also argues that road safety professionals and governments need to be more
questioning of the worth of educational and training approaches and have the courage to
say ‘NO’ to advocates, lobbyists and politicians who want to expend funds and
resources on unproven programs (Christie, 2002). Christie’s argument however is based
on the major functions of public health evaluation. Such initiatives, programs and
interventions rely on a reduction in death and injury, being cost-effective in terms of
expenditure of funds and resources, and do no harm (MacIntyre, 2000, Turnock, 1997).
The argument against education programs also encompasses the short-term
effectiveness. Additionally, Grossman and Garcia (1999) found in their study that
66
examined systematic reviews of health promotion programs to increase seatbelt wearing
among young children, that the benefits were short-lived. This is one argument that must
be recognised in the development of any type of road safety intervention that is
educational in nature.
Review of the literature has found that most evaluation studies or systematic
reviews have been conducted on educational campaigns or long-running programs.
Very little research has focused on theoretically developed interventions that are brief in
nature. Although anecdotal evidence would indicate that such interventions would likely
to be ineffective, no research in level crossing safety has ever been conducted in the
delivery of educational interventions per se. Therefore, efforts to intervene on specific
road user groups through a single intervention educational approach deserve
consideration. An appraisal of level crossing safety programs both in Australia and
overseas is provided below.
2.7.5 Level crossing safety programs and interventions
2.7.5.1 Overview
Despite research that suggests an informed approach offers a more efficient and
less costly process with increased probability of success (Witte, 2000), it appears that the
development and analysis of effective messages to promote safe driving behaviour is
limited in the rail safety literature. In reviewing the literature on educational programs
for motorist safety, it must be noted that the context of a level crossing educational
program or intervention would be vastly different from generic road safety education
programs or interventions. The reason for this is that a level crossing program and/or
intervention would target specific issues pertaining to crossing safety rather than
providing general educational information.
2.7.5.2 Australian interventions and programs
To date, there has been no scientifically evaluated level crossing education
program or interventions for motorists in Australia. However, there have been
67
numerous attempts by many jurisdictions to try to ‘educate’ the public about both
pedestrian and motorist safety at level crossings.
In South Australia, the ‘Don’t Play with Trains’ safety campaign (2004/05 –
2006/07) aims to inform all road users that their behaviour is the key factor in collisions
at level crossings. The television and radio commercials were first aired in June 2005
and again in July 2006 during National Rail Safety Week (23-29 July 2006). The
campaign was also aired in February 2007. Two commercials were developed (one
commercial for motorists and the other for pedestrians) with the aim to demonstrate that
risk taking at level crossings for both motorists and pedestrians is hazardous
(Department for Transport Energy and Infrastructure, 2006). The commercial for
motorists focuses on two intertwining scenarios. The first scenario seen is that of two
young men driving a work utility vehicle in a rural/remote area whilst talking about
football. The utility comes across an unprotected level crossing and the driver does not
stop at the crossing as the commercial leads the viewer to believe that the driver assumes
there is no train coming. The viewer observes the utility being hit by the train with the
viewer believing it is a fatal collision. The other scenario is that of a middle-aged mother
with two primary school-aged children in the back of the car. She is talking to the
children about what they did at school when she enters a metropolitan level crossing
behind a truck and gets caught when the boom gates are descending. The view also
observes this car being hit by the train with the viewer believing that it is a fatal
collision. An audio script is read by a male while both scenarios are played out. At the
end of the commercial the voiceover says “don’t play with trains”.
The impetus for this campaign was brought about by the collision between a
passenger train and a bus and car on the level crossing at Park Terrace, Salisbury in
South Australia in October 2002. Tragically, four people died (including the bus driver)
and 26 people sustained serious injuries. The investigation into this collision resulted in
a number of recommendations including the development of a public education
campaign highlighting the dangers of level crossings.
The NSW Roads and Traffic Authority (RTA), with assistance from the Railway
Infrastructure Corporation, conducted research into motorist behaviour and collisions at
level crossings. This market research found that rural/remote community attitudes to
68
level crossings included complacency and high risk behaviour. The objectives of this
campaign were to:
• Increase awareness of the importance of obeying road rules at level crossings;
• Create awareness of the dangers caused by the stopping limitations of trains; and
• Increase safe behaviours by motorists using railway level crossings.
(NSW Roads and Traffic Authority (RTA), 2006)
In Victoria, a $1 million advertising campaign was launched in November 2005,
including television, radio and outdoor advertising. This campaign was prompted by the
number of traffic offences at the Springvale Road level crossing that was part of a video
camera trial. More than 5000 offences were recorded during 28 days of recording. A
combination approach is being used in Victoria, with trialing of new road safety
initiatives as well as greater enforcement at six of the highest risk level crossings in
Melbourne. The first phase of the education campaign targeted metropolitan motorists
with the campaign ‘Don’t risk it. Always keep the crossing clear’, while in regional
Victoria, the campaign is ‘Don’t risk it. Slow down and be prepared to stop’. Other
initiatives that are being trialed as part of this campaign include:
• A trial of the effectiveness of yellow cross-hatching (yellow grid lines painted on
the road) at particular level crossings to prevent queuing;
• A review of the operation of traffic signals located near crossings to ensure they
do not trap vehicles on crossings;
• Trialing new illuminated signs to warn pedestrians of the approach of second
trains; and
• A review of the safety features of Victoria’s 3,000 crossings to determine the
most appropriate treatment for each.
(Metlink Victoria, 2005)
The next phase of this education campaign, developed with VicRoads, Transport
Accident Commission (TAC), Connex, V/Line and Victoria Police, aims to also target
69
high risk pedestrians including children, teenagers, the elderly, cyclists and the mobility
impaired.
In Queensland, the ‘Rail Smart’ campaign was launched in July 2006 as an
initiative that brings together a number of safety related activities that QR conducts on
an ongoing basis, including safety awareness and education activities which addresses
level crossing safety, corridor trespassing and general rail user safety issues. This
campaign is broadcast through a 30 second television commercial, which has two main
themes: (1) QR Rail Smart (generic rail safety message) and (2) specific safety messages
such as ‘be alert near level crossings’. This campaign is also delivered through:
• Television Community Service Announcements;
• Supporting collateral;
• Rail property media placements;
• Development of a Rail Smart message in conjunction with the Queensland Police
Service and the Queensland Police Citizens Youth Club;
• Incorporating Rail Smart activities in the QR Reds Talent ID Clinics throughout
regional and metropolitan Queensland during the 2007 Super14 Season; and
• Delivering Rail Smart messages through QR's community education unit (CEU)
as part of its state-wide primary, secondary and tertiary student programs. (In
2005 160,659 students were addressed in 500 schools in Qld).
(Queensland Rail, 2007)
2.7.5.3 International programs
‘Operation Lifesaver’ is probably the most well-known education and awareness
program for level crossing safety in the world. This program is currently run in the
United States and Canada as a non-profit program, with its mission to educate both
drivers and pedestrians to make safe decisions at level crossings. The program began in
1972 as a one-time, six-week only public awareness campaign in the state of Idaho
(Hall, 2002). This campaign focused on several key audiences, with volunteer presenters
delivering messages to driver education students, professional truck drivers, school bus
drivers and law enforcement agencies. During this six-week program in Idaho, a
reduction in fatalities was observed. The following year, the program was rolled-out
70
into other states, with lowering of collision rates observed. By 1986, there were 49
independent Operation Lifesaver programs operating in the United States. These
programs focus on the three E’s of road safety: engineering, enforcement and education.
A study by Savage (2006) which used a negative binomial regression to estimate
whether variations in Operation Lifesaver activity are related to the number of collisions
and fatalities at level crossings. Annual data on the experiences of 46 states from 1996-
2002 were used, as only data for this period was able to be analysed due to incomplete
and questionable data from previous years. Additionally, the effect on the number of
fatalities cannot be concluded with statistical certainty, due to a small number of
fatalities compared with collisions. Savage (2006) suggests that there is considerable
year-to-year random variability in the number of fatalities, hence making it more
difficult to find statistical robust relationships. Savage (2006) also suggests that
quantifying and evaluating enforcement activities is more difficult than quantifying the
effects of engineering solutions. Although Savage’s (2006) study only used annual data
from 1996-2002, due to being the only complete and accurate data since 1972, an earlier
study by Mok and Savage (2005) analysed pooled data from 1975-2001. This study
disaggregates the improvement in level crossing collisions and fatalities into its
constituent causes. Negative binomial regressions were also used in this data analysis.
The findings from this study suggest that approximately:
• Two-fifths (40%) of the decrease in collisions and fatalities are due to factors
such as reduced drunk driving and improved emergency medical response;
• One-fifth (20%) of the decrease in collisions and fatalities is due to the
installation of gates and/or flashing lights;
• One-tenth (10%) of the decrease in collisions and fatalities is due is due to
closure of crossings resulting from line abandonments or consolidation of
infrequently used crossings;
• One-seventh (14%) of the decrease in collisions and fatalities is due is due to
‘Operation Lifesaver’; and
• One-seventh (14%) of the decrease in collisions and fatalities is due is due to the
installation of additional lights on locomotives (during the mid 1990’s).
(Mok and Savage, 2005)
71
2.7.5 Limitations of current level crossing safety programs
Reviewing the evidence of the effectiveness of road safety campaigns reveals that
many have typically been developed in a haphazard manner, and may have limited
effectiveness in improving road safety. Mass media advertising, such as those used by
all jurisdictions in Australia to tackle level crossing safety and the ‘Operation Lifesaver’
program in the United States and Canada, have neither been evaluated for effectiveness
in terms of self-reported or intended behaviour. Additionally, in Australia campaigns
have typically been isolated (i.e. not involving increased police presence) and therefore
have not attempted to apply principles for maximising the effectiveness of road safety
campaigns. It is noted that observational studies or recordings is an expensive road
safety initiative, and is most likely economically not viable for all level crossings in any
country. However, to date there has been no effort or attempt to develop any type of
campaigns on theory.
It is well known that theoretically grounded campaigns developed in accordance
with research and targeting specific road safety issues can provide a more effective
means of risk management. Indeed research investigating the effectiveness of
educational programs targeting specific road safety issues has found programs to be
highly effective in reducing road crashes (Cameron, 1993, Durkin, 1999, Powles, 1993).
Finally, evaluations of cost/benefit ratios by Guria (1999) of road safety initiatives
programs, suggests that education and advertising campaigns produce high incremental
returns as compared to alternative methods of risk management such as engineering
approaches. These positive research outcomes coupled with this favorable economic
evaluation, suggests that the current investment in road safety programs is below optimal
and warrants further research into theoretical models and evaluative research appropriate
for effective program development.
With engineering approaches at level crossings no longer being a viable option in
terms of cost, the role of targeted educational interventions in informing future programs
needs to be examined. Without the evaluation of targeted interventions for specific road
user groups, campaigns and programs generated by government authorities will no doubt
be developed without the application of theory or the findings from scientific evidence.
72
2.8 HUMAN FACTORS CONTRIBUTING TO COLLISIONS
2.8.1 Overview of accident causation
There are two perspectives on human factors in accident causation: the ‘old view’
and the ‘new view’ (Dekker, 2002). The ‘old view’ (Reason, 2000, American Medical
Association., 1998) sees human error as a cause of failure by which it is seen in the
following way:
• Human error is the cause of most accidents;
• The engineered systems in which people work are made to be basically safe;
their success is intrinsic. The chief threat to safety comes from the inherent
unreliability of people; and
• Progress in safety can be made by protection these systems from unreliable
humans through selection, procedurelisation, automation, training and discipline.
(Dekker, 2002, p372)
The ‘new view’ sees human error not as a cause but as a symptom of failure
(American Medical Association., 1998, Hoffman, 2000, Reason, 2000, Woods, 1994)
and includes the views:
• Human error is a symptom of trouble deeper inside the system;
• Safety is not inherent in systems. The systems themselves are contradictions
between multiple goals that people must pursue simultaneously. People have to
create safety;
• Human error is systematically connected to features of people tools, tasks and
operating environment. Progress on safety comes from understanding and
influencing these connections; and
• Human error represents a substantial movement across the fields of human
factors and organisational safety and encourages the investigation of factors that
easily disappear behind the label ‘human error’.
(Dekker, 2002, p372)
73
Recent studies have found that unsafe driving acts can be classified into three
distinct categories: errors, lapses and violations (Reason, 1990b). According to Reason
et al. (1990b), in considering the human contribution to accidents, it is necessary to
make a distinction between these categories. Previous research has indicated that these
three main types of aberrant driver behaviour have different psychological origins and
demand different modes of remediation (Parker, 2000). According to Parker, Reason,
Manstead and Stradling (1995):
• Lapses are typically absent-minded behaviours with consequences mainly for the
perpetrator, posing no threat to other road users;
• Errors are typically misjudgments and failures of observation that may be
hazardous to others; and
• Violations involved deliberate contraventions of safe driving practice.
Such measures of driving behaviour are typically examined within the Driver
Behaviour Questionnaire (DBQ) (Reason, 1990b). There have been several versions of
the DBQ over the years, with Lawton et al’s (1997a) study subsequently extending the
violations scale by added new items to the original DBQ. Factor analysis conducted by
Lawton et al (1997a) showed that violations can be split into two distinctive scales
according to the reason why drivers violate. The aggressive violation scale contains an
interpersonally aggressive component, while the ordinary violation scale contains
deliberate deviations from safe driving without a explicitly aggressive aim (Lajunen,
2004b). Violations rather than errors or lapses have been found in the literature to be
related to road crash involvement (Lajunen, 2004b, Parker, 2001, Reason, 1990b),
however research with elderly drivers has shown relatively high error and lapses scores
in predicting road crash involvement (Parker, 2000). It has been shown that errors,
lapses and violations are three empirically distinct classes of behaviour, however it is
difficult to identify such behaviours without the reconstruction of the sequence of events
in any particular crash (or collision).
“Reconstructing the human contribution to a sequence of events that led up to an
accident is not easy” (Dekker, 2002, p.373). Accident investigators and coroners are
rarely there when events unfold around the people under investigation, and as a result
74
their actions and assessments may appear not only controversial, but in fact perplexing
when seen from a different point of view (Dekker, 2002). Investigators and safety
authorities, thanks to hindsight, consider that they know more about the incident or
accident than the people involved (Dekker, 2002). Dekker (2002) argues that hindsight:
• Means being able to look back, from the outside, on a sequence of events that led
to an outcome that has already happened;
• Allows almost unlimited access to the true nature of the situation that surrounded
people at the time (where they actually were versus where they thought they
were; what state their system was in versus what they thought it was in); and
• Allows investigators to pinpoint what people missed and should not have missed;
what they did not do but should have done.
(p373)
From this hindsight perspective, the entire sequence of events can be exposed.
However, according to Dekker (2002) this contrasts fundamentally with the point of
view of the people who were inside the situation as it unfolded around them. To these
insiders, the outcome was unknown, nor the entirety of surrounding circumstances,
although they contributed to the course of the sequence of events on the basis of what
they understood to be an evolving situation (Dekker, 2002). For the majority of
accidents in complex situations, Dekker (2002) argues that people were doing exactly
the sorts of things they would usually be doing – the things that usually lead to success
and safety. Wigglesworth’s (2001) study of detailed police reports prepared for the
coroner in level crossings fatalities, supports this notion in one sense. This study
concluded that in most cases level crossing collisions occurred to a law-abiding citizen
going about his or her daily work and was attributed to human overload rather than any
breach of regulation (Wigglesworth, 2001). Reason (1997), Sanne (1999) and Woods et
al (1994) argue that mishaps are more usually the result of everyday influences on
decision making than isolated cases of erratic individuals behaving unrepresentatively.
Dekker (2002) argues that people are doing what makes sense to them given the
situational indications, operational pressures, and organisational norms presented to
them at the time. To put it simply, people do not want to be involved in a road crash and
75
they are generally doing what makes sense to them at the time (Dekker, 2002). Such
behaviour is the result of errors and lapses only, and relies on the different concepts of
performance and behaviour.
Performance and behaviour are two fundamental components of driving that are
inextricably intertwined and need to be recognised in both the development of road
safety countermeasures and factors contributing to road crashes. Performance relates to
“our abilities to perceive and react to circumstances in an appropriate and- timely
manner. It is the manifestation of what we commonly refer to as ‘skill’ (Henderson,
1991). Behaviour refers to what a driver actually does on the road, not what they are
capable of doing. According to Henderson (1991), behaviour “embraces not only
performance as in fact performed, but also attitudes to the task, cultural differences and
pressures, and the way we perceive and respond to risk” (p4). There is great difficulty
however for road safety researchers in investigating behaviour, as information is harder
to come by and is usually only able to be examined in advanced laboratory conditions
(Henderson, 1991).
It must be noted that the act of driving is taken for granted by most motorists,
however it is a complex task that is dependent on a variety of cognitive and psychomotor
performance abilities to be intact, such as alertness, attention, multitasking, memory,
coordination, and visual spatial perception (Moller, 2004). According to Ogden (1996),
driving can be described by three essential tasks: navigation, guidance and control.
These three tasks require the driver to receive inputs from a driving environment,
process them, make predictions about alternative actions, decide which are the most
appropriate, execute the actions, observe their effects through feedback and process new
information (Lay, 1990). Wang (2002) suggests that “essential in performing these tasks
is a driver’s ability to make relatively accurate estimates of the safety of the driving
environment” (p253).
With regards to the view taken to explain accident causation at level crossings,
there appears to be two schools of thought. According to research conducted by the
University of Calgary, the study of human factors of level crossing collisions attempts to
place human error in the context of perceptual, memory, cognitive and motor
76
capabilities (Caird, 2002). Figure 5 illustrates factors contributing to vehicle-train
collisions at level crossings (Caird, 2002).
Figure 5: Contributing factors in highway-railway grade crossing collisions.
However, according to one Australian expert in level crossing safety, the human
factor approach reviews the system as a whole and does not apportion blame, but rather
seeks to observe patterns of human behaviour that differ from designer expectations and
consequently to gain an understanding of the human limitations that have resulted in
such differences (Wigglesworth, 1978, Wigglesworth, 2001). This specifically relates to
passive level crossings in that the designer expectation is that motorists will look in both
directions to see if a train is approaching, at a speed that enables them to stop if
necessary (Wigglesworth, 2001, Leibowitz, 1985, Richards, 1990, Russell, 1993).
Australian coroners have tended to agree with this view of Wigglesworth by reporting
that from a systematic point of view, level crossing collisions usually have more than
one contributing factor (i.e. being the non-observance of the Australian Road Rules).
Some coroners have indicated that there are numerous causal factors relating to:
• Road design (the number of entry/exit points);
77
• Road traffic lights and the inter-link with the level crossing warning system;
• The width of the crossing;
• Probable lack of awareness by road vehicle drivers of the road traffic rules as
they relate to level crossings;
• The lack of ‘near hit safety’ reporting at level crossings; and
• The lack of a focused body to oversight and undertake risk based assessments of
level crossing safety.
(Australian Transport Safety Bureau, 2002a, pvi)
2.8.2 Driver interaction with different protection systems
2.8.2.1 Active level crossings
Drivers are notified when approaching an active (automated) level crossing of a
train’s presence or an approaching train through the use of flashing lights and/or (visual)
and bells (aural). These warning systems are activated by the train circuitry for a period
of approximately thirty seconds before the arrival of the train. Such systems provide a
much better assessment of the likelihood of an oncoming train than passive level
crossings (stop or give-way signs only). Warning and protection systems at actively
protected level crossings (flashing lights and/or boom gates) have been found to have a
very high level of reliability. Stott (1987) in his research found that collisions occurred
as a result of the failure of the motorist to pay attention to the warning systems at
crossings rather than a mechanical failure of the warning system.
Actively protected crossings usually bring about a reduction in delay to the road
user, but also places responsibility on the motorist to observe and obey the road traffic
signals. Generally, driver observance at active level crossings is of a high standard
(Pickett, 1996) but violations do occur, which may or may not lead to a collision.
According to Pickett and Grayson (1996) there are three categories of drivers that have
been identified as likely to be involved in an accident at actively protected level
crossings:
• Those who are unwilling to stop because they believe they have plenty of time to
cross before the train arrives;
78
• Those who are unable to stop because they are too close to the tracks at the onset
of flashing lights or because someone is driving too close behind; and
• Those who are unaware of the signals because they are inattentive or distracted.
This list of categories suggested by Pickett and Grayson (1996) is somewhat
incomplete as it does not include those drivers that lack knowledge and awareness about
warning and/or protection systems or those drivers that find these systems complex and
confusing. Additionally, it doesn’t include drivers who are familiar with a particular
level crossing. Wigglesworth’s study (1979) of 85 consecutive fatal collisions involving
motor vehicles and trains at all types of level crossings in Victoria, found from detailed
police reports prepared for the coroner that most fatalities were not linked with breach of
regulations, but instead pointed to the presence of stressors such as family bereavement,
financial or employment problems, or distraction of the driver by events inside or
outside the car. This study concluded that in most cases, the accident occurred to a law-
abiding citizen who was not in breach of any regulations, but rather going about their
daily work (Wigglesworth, 1979). Familiarity was found to be linked with at least 86%
of these fatalities as drivers lived locally and were familiar with the existence of the
crossing (Wigglesworth, 1979).
2.8.2.2 Passive level crossings
Passive crossings usually have extremely low volumes of both rail and road traffic
and are traveled over by the same small cohort of local drivers who may or may not be
representative of the driving population generally (Wigglesworth, 2001). Although the
assumption by the public and some researchers is to upgrade passive crossings to active
protection systems to make them safer, this has been rebutted as active protection
systems continue to be linked with collisions. In August 2004, 3 people were fatally
injured when their car was hit by a train while queuing over the St. Albans level crossing
in Melbourne which was actively protected with both flashing lights and boom barriers.
Therefore, it can be seen that even actively protected crossings may also be an
environment for error-producing decisions.
79
2.8.3 Perception of risk
Risk perception is the subjective assessment of the probability of a specified type
of accident happening and how concerned we are with the consequences. To
perceive risk includes evaluations of the probability as well as the consequences of
a negative outcome.
(Sjöberg et al., 2004, p8)
There is an extensive body of literature devoted to the psychological work on the
subject of perceived risk. However, it is generally unknown how and why individuals
differ in their judgment of risk (Kraus and Slovic, 1988). The perception of driving risk
is reported to be subject to systematic distortions (Rafaely et al., 2006). This is
particularly evident in the younger driving population (i.e. under 25 years of age).
Perceived risk (a driver’s feelings about the level of risk for a particular driving
situation) has been demonstrated to be lower for younger drivers than for older drivers.
Younger drivers typically underestimate the risk of being involved in a road crash
(Deery, 1999, Dejoy, 1992), linking them to risk taking behaviour (Glendon et al., 1996)
as well as risky driving behaviour (Deery, 1999, Harre, 2000). Gardner and Steinberg’s
(2005) simulated driving study found that the presence of peers as passengers influenced
risk decision-making and perceived risk. Compared to when the driver was alone,
participants reportedly took more risks, focused more on the benefits than the possible
costs of risky driving behaviour, and make riskier decisions (Gardner and Steinberg,
2005). Additionally, some research has indicated that young drivers have been shown to
perceive their own risk as lower than that of their peers (Finn and Bragg, 1986).
Compared with the younger driver population, older drivers have been found to perceive
their driving ability better than or equal to that of their peers (Holland, 1993, Freund et
al., 2005), and better than their younger counterparts (i.e. younger drivers aged under 25
years) (Groeger and Brown, 1989).
Recent studies in Canada reveal that motorists view level crossings as a low risk
safety problem (Beirness et al., 2003). This research found that when “examined in the
context of other road safety issues, highway/railway crossing safety is viewed as being
amongst the least serious” (p9). The survey was administered by telephone and 1209
80
drivers agreed to participate. Figure 6 shows the average ratings provided by the
random sample of Canadian drivers (Beirness et al., 2003). A scale of 1 representing
‘not a problem at all’ to 6 representing ‘an extremely serious problem’ was used. As
can be seen, many survey participants believe that the most serious road safety issues are
drink driving and red light running (average rating of 5 or greater), while
highway/railway crossing safety received an average rating of 3.0. This rating was also
below safety issues of vehicle defects and distracted drivers (Beirness et al., 2003). This
has been the only known research that has included level crossing driving with other
driving behaviours.
Figure 6: Perceived risk ratings by Canadian drivers
2.8.4 High risk behaviours and risk taking
To understand risk taking at a deeper level, it is important to understand the
cognitive and perceptual processes underlying driving behaviour in an environment that
has potential traffic hazards. Deery (1999) conducted an extensive review of hazard and
risk perception among novice drivers. He recommends that “before defining risk and
hazard perception, it is important to note that they are concepts that reflect, at least in
81
part, drivers’ subjective experiences and thus should be distinguished from objective
risk” (p226). The subjective experience of risk in potential traffic hazards, is referred to
as risk perception (Deery, 1999). Brown and Groeger (1988) propose that there are two
inputs to determine risk perception:
• Information regarding the potential hazards in the traffic environment; and
• Information on the ability of the driver (and the capabilities of the vehicle) to
prevent those potential hazards from being transformed into actual accidents.
Figure 7 illustrates the domain of hazard perception such as information about
potential traffic hazards and the process of identifying hazardous objects and events, and
judging and quantifying their potential to be dangerous (Armsby et al., 1989, Brown,
1988). This model also illustrates driver’s beliefs about their ability in handling
hazardous situations, which is the result of their self-assessed driving ability (Deery,
1999). Hazard perception, using this model, involves elements of both driving skill and
subjective experience. In comparison to risk perception, risk acceptance is the degree to
which a driver is willing to accept risk, with the driver having a level of perceived risk
or risk threshold (Deery, 1999). Deery (1999) suggests that because “driving is
essentially a self-paced activity, a driver determines the difficulty of his/her task by
setting and accepting different risk thresholds” (p226). He proposes that a driver’s
motivation determines their risk threshold for behaviour alternatives, with there being
several reasons why a driver may take a risk including:
• Driver may exhibit poor risk perception – e.g. the driver may perceive low levels
of risk in maintaining a relatively high driving speed. Poor risk perception may
result from a driver misjudging the distance to a level crossing or the braking
distance of their vehicle. It may also occur from the driver over-estimating their
ability to deal with a problem should it arise;
• Driver may possess relatively poor levels of driving skill – e.g. the time taken to
detect a level crossing and make a decision to slow down may be excessive,
thereby delaying the response time to release the accelerator or brake; and
• Driver may have a high level of risk acceptance – e.g. they may correctly
perceive the risk of driving over a passive level crossing without scanning for
82
trains, and may also possess a relatively high level of driving skill, but they may
have chosen to accept the risk to minimise time delays.
(Deery, 1999, p228)
Additional to the performance of high risk behaviours by drivers, research has
indicated that risk taking has been linked to the involvement in road crashes. However,
only a handful of studies have actually measured the association between injury
resulting from a road crash and risk taking behaviour (Bell, 2000, Crutcher et al., 1994,
Perneger, 1991, Rajalin, 1994, Turner, 2004). Risky driving, in the broad sense,
encompasses behaviours such as drink driving, drug driving, speeding, red light running,
tailgating, weaving in and out of traffic, unsafe overtaking and distracted driving
(Beirness and Simpson, 1988, Jessor, 1977, Jonah and Dawson, 1987). When drivers
ignore the warning signals and signs at level crossings, often in an attempt to ‘beat the
train’, this act can be interpreted as risk taking (Reason, 1997). However, research
indicates that the cause of fatal collisions at level crossings is largely not the result of
deliberate violations of the road rules, but rather inattention (Wigglesworth, 1979).
83
Figure 7: Driving behaviour responses to potential hazards
84
Risk acceptance has also been examined by Wilde (Wilde, 1982, Wilde, 2001).
Wilde developed of theory of ‘risk homeostasis’ in which individuals are prepared to
accept a given target level of risk, resulting from a gap between the perceived level of
risk and an (exogenously given) target level of risk (Gossner and Picard, 2005). Other
researchers argue that the ‘risk homeostasis’ theory is possible only when under very
restrictive conditions (Janssen and Tenkink, 1988). Janssen and Tenkink (1988)
argue that risk compensation can occur under more general conditions: that is, the
safety effects anticipated from engineering measures will be discounted to a large
degree by shifts in behaviour. This is supported by conclusions from an OECD report
(1990) that indicates that behavioural adaptation or compensation does not eliminate
the safety gains obtained, but it may reduce the effectiveness of road safety programs
in a number of cases (Gossner and Picard, 2005). Given such empirical evidence that
motorists adapt or compensate their behaviour to changes in their driving environment
(such as road safety engineering and traffic countermeasures), it would be reasonable
to assume that motorists would drive more safely at high risk intersections such as
level crossings. However, the evidence suggests otherwise.
At level crossings, it is important to recognise that within the sub-set of
motorists whom engage in high risk behaviours, there are two main types of
individuals. As discussed previously, some individuals purposively and knowingly
engage in risky behaviours such as trying to ‘beat the train’ for the thrill of it or
possibly because of impatience at having to wait for the train to pass. When motorists
ignore the warning signals and signs at level crossings, often in an attempt to ‘beat the
train’, this act can be interpreted as risk taking (Reason, 1997). Willful violations
include driving around the gates and driving around cars stopped at the crossing
(Abraham et al., 1998, Berg, 1981, Meeker, 1997). Results from Berg et al (1981) and
Abraham et al (1998) suggest that the timing of crossing signals as a reason motorists
crossed when flashing lights were still activated. In Berg et al’s (1981) study, drivers
were more likely to cross after the signal had been activated for more than 30 seconds.
Unnecessary waiting times to the presence of slower moving trains (e.g. freight) on
track circuits designed to accommodate faster trains (e.g. passenger), are identified as
a reason motorists violate warnings signals and signs (Berg, 1981). Respondents to
Abraham et al’s (1998) study indicated that they violated warning signals or
protection systems because the train was not in sight or because it had been stopped
85
for an unreasonable amount of time. Such assumptions by drivers are inherently
dangerous. Caird et al (2002) report that:
• Restricted visibility may prevent motorists approaching flashing light
crossings from seeing the train and they therefore erroneously assume that a
train is not present;
• If a train is stopped near a crossing with multiple tracks, the signals observed
by the motorist may actually be for another train approaching from either the
opposite direction or from behind the stopped train on a parallel track. These
types of collisions are called second train collisions; and
• Motorists’ judgments of how far away a train is from a crossing may be
affected by perceptual factors, such as looming.
In order to further understand the types of individuals who are most at risk for
vehicle-train collisions, Witte and Donohue (2000) surveyed 891 randomly selected
residents from the U.S.A. These researchers found that whilst the majority of
respondents perform safe driving behaviours, a sub-group of 10-20% labeled ‘risk
seekers’ engaged in extremely risky behaviours (Witte, 2000). More specifically,
approximately 10% of the respondents perceived trying to ‘beat the train’ across the
tracks to be exciting, and almost 14% reported that they would drive around the gates
with the flashing lights even if a train was in sight. Analyses revealed that members
of this ‘risk seeker’ sub-group were disproportionately male, with a high propensity to
engage in other risky behaviours such as smoking, drinking alcohol and fighting
(Witte, 2000). These sub-group members were more likely to have experienced
personal frustration at level crossings and to exhibit biased judgment processes
regarding their ability to ‘beat the train’. Additionally, they also reported high
‘sensation seeking’ tendencies to engage in novel experiences and to minimise
boredom. The researchers suggest that these strong sensation seeking tendencies
predispose this sub-group to experience greater frustration and demonstrate higher
judgment distortions, which consequently leads to riskier behaviour.
Other individuals inadvertently put themselves and others at risk through failing
to successfully recognise a dangerous situation (Wigglesworth, 1979, Abraham et al.,
1998, ARRB Transport Research., 2002, Berg, 1982, Caird, 2002, Leibowitz, 1985,
Meeker, 1997, Pickett, 1996, Shinar, 1982). These motorists are engaging in risky
behaviours due to a lack of awareness of approaching trains or an inability to judge a
86
safe distance to cross in front of an oncoming train. Unintentional violators, who are
unable to stop at a level crossing for a variety of reasons (such as speeding or being
too close to the intersection at the onset of the flashing lights without causing danger
to other road users), have made a conscious decision to cross the intersection and
generally feel that circumstances leave them with no choice but to violate the road
rules.
2.8.5 Familiarity
Familiarity with a level crossing may influence a motorist’s behaviour in
numerous ways (Caird, 2002). “Crossing familiarity and an expectation that a train
will not be present have the potential to lull drivers into complacency or poor looking
habits” (Caird, 2002, p101). Motorists that expect long delays may violate traffic
signals and possibly reduce their scanning behaviour to detect a train’s approach
(Abraham et al., 1998). Abraham et al’s (1998) study revealed that 87% of motorists
who committed a violation at an active crossing used the crossing regularly. This
finding suggests that familiarity may encourage motorists to take greater risks
(Abraham et al., 1998). However, determining an accurate picture of a motorist’s
familiarity with a level crossing is difficult as fatally injured motorists are unable to
testify (Wigglesworth, 1979). To date, no data is recorded by the ATSB on familiarity
of motorists.
In case study investigations (Wigglesworth, 1979, National Transportation
Safety Board, 1986), findings have indicated that motorists involved in crossing
collisions were familiar with the crossing. Wigglesworth’s (1979) study of 85
consecutive fatalities in Victoria, Australia between 1973 and 1977 found that 73 of
the 85 motorists (86%) were considered to be familiar with the crossings at which
their collision occurred. These motorists lived within one mile of the level crossing,
classifying them as familiar with the crossing. However, motorists who did not live
close by to the crossing but were familiar with it due to employment or other reasons,
could actually be familiar with the crossing (Wigglesworth, 1979). Additionally,
motorists who lived close by to the crossing may rarely drive over it as there is no
need to. The National Transportation Safety Bureau’s (NTSB) (1986) study found
similar results to Wigglesworth’s (1979) study. The NTSB study estimated that
nearly 85% of motorists involved in a fatal collision were familiar with the crossing.
87
However, this study did not indicate the classification of familiarity unlike
Wigglesworth’ study.
Overall, it is highly likely that motorists who are familiar with particular
crossings may become complacent and take greater risks than whilst driving at
unfamiliar crossings. According to Pickett (1996) motorists may “also base their
decision to cross on their previous experience of either the same crossing (familiarity)
or of other crossings (association). The concept of mental set says that when people
are exposed to the same phenomenon repeatedly, they come to expect it” (p5). Pickett
(1996) also suggests another situation, with motorists familiar with one crossing,
transferring their experience to a new crossing and not being as vigilant in the new
situation.
2.8.6 Approach behaviour
Once a motorist has recognised the presence of a crossing, it is the
responsibility of the motorist to also determine whether an approaching train is
imminent (Ward and Wilde, 1995). However, according to Ward and Wilde (1996)
motorists approaching crossings may be uncertain about the probability of
encountering a train, particularly at passive crossings. Ward and Wilde (1996)
suggest that “uncertainty will be high without reliable knowledge of the scheduling of
train arrivals based on past experience, either because (i) the trains at the site do not
adhere to a regular schedule, or (ii) there is no prior experience with the site” (p63).
‘Decisional uncertainty’ may lead to inconsistent behaviour within and between
motorists, particularly when a decision to stop at a level crossing is not feasible (e.g.
traveling to fast to stop safely) or not warranted (e.g. when a train has stopped a long
distance from crossing) (Ward and Wilde, 1995, Ehrlich, 1989). A motorist must
determine how to respond at level crossings, based upon joint consideration of their
own approach parameters and those of the train (i.e. direction of approach, estimation
of speed and distance from crossing) (Meeker, 1989).
The approach parameters of a motorist include the restriction of lateral sight
distance. When there is limited sight distance to dense vegetation or approach angles,
it would be logical to assume that such restrictions are associated with greater
incidence of collisions because of the greater hazard associated with the obstruction of
lateral visibility (Ward and Wilde, 1996). Nevertheless, there has been no consistent
88
demonstration of any relationship between lateral sight distance and collision history
(Russell, 1974a, Zalinger et al., 1977, van Belle et al., 1975). Ward and Wilde (1996)
offer an explanation for this phenomenon based on the concept of risk homeostasis.
They suggest that motorists compensate their behaviour (such as speed reduction) in
response to the perceived risk associated with restricted visibility, particularly at
passive crossings (Wilde et al., 1987, Ward and Wilde, 1996). Such behaviour,
proposed by Ward and Wilde (1996) is expected to maintain a ‘more-or-less’ constant
safety margin. This notion has been supported by their study in examining the effects
of enhancing lateral sight distances at a passive crossing in Ontario, Canada. The
study’s methodology also included parallel observations with an untreated site to
control for secular confounding (Ward and Wilde, 1996). The findings of their
research met their expectations – “improvement at lateral sight distances resulted in
an upstream shift toward longer search durations and a tendency towards faster
approach speeds, but failed to produce a calculated net safety benefit” (p63).
Additionally, it was found that enhancement treatment at the treatment site reduced
perceived risk by local motorists (Ward and Wilde, 1996).
As can be seen, there has been little extensive research published on the safety
issue of approach behaviour at level crossings. Other than a few well-known
published studies in the area, little recent research is available. This is an area greatly
lacking in level crossing safety and requires further investigation, particularly in
Australia.
2.8.7 Poor knowledge of road rules
According to Donohue, Wendelken, Crone and Bunge (2005), “a detailed
account of how we use rules to make decisions would constitute an important advance
in our understanding of human behavior” (p1140). However, no research to date has
focused on knowledge of road rules and level crossing driving behaviour, although
anecdotal evidence suggests that many Australian drivers have a poor knowledge of
road rules at level crossings.
89
2.8.8 Distraction
Driver distraction is a serious problem in road safety and has long been
suggested to be a contributing factor to collisions (Treat, 1977). Although there is a
considerable body of evidence to support the notion that motorists can become
distracted while they drive, until recently little attention has been paid to distraction as
an important risk factor for road crash involvement. Although the number of road
crashes in Australia for which distraction is a contributing factor is unknown,
international evidence indicates that distraction is a significant contributing factor to
road trauma (Regan, 2003). National data in Australia has in the past not recorded
distraction in a detailed manner, but as Regan and Young (2003) suggest “although
the full extent to which distraction is a causal factor in road crashes in Australia is not
yet known, there is converging evidence that it is likely to be a significant problem
here” (p1).
While Lam (2002) proposes that distraction comprises a variety of activities,
situations and circumstances that a motorist may be exposed to while performing a
driving task, Ranney et al (2000) argue that it may be characterised as any activity
that takes a driver’s attention away from the task of driving. The vast majority of
driving is characterised by activities involving straightforward control of the vehicle,
such as sustaining appropriate speed, headway, and lane position within the road
environment. These familiar activities require very little cognitive processing and
motorists are often able to engage in other cognitive activities simultaneously without
noticeable detriment to the driving task (Ranney, 2000). Brown (2003) suggests
however that safe driving of a motor vehicle requires a wide range of skills and
abilities, with driving performance being sub-optimal when the driver is distracted,
either from within or external to the vehicle (Brown, 2003). Cognitive overload and
cognitive underload are the two elements that are identified as being fundamental to
motorist distraction (Wallace, 2003). Researchers generally agree that distraction can
take place even when a driver is concentrating on the task at hand.
The National Highway Traffic Safety Administration (NHTSA) proposes that
there are four distinct, although not mutually exclusive, forms of motorist distraction.
These include: visual, auditory, biomechanical (physical) and cognitive (Royal,
2002). While Regan and Young (2003) have classified motorist distraction in their
review of the literature as being either technology based distraction or non-technology
90
based distraction, with some overlap occurring. While other researchers suggests that
internal-to-vehicle and external-to-vehicle are the two major classifications of
distraction (Wallace, 2003, Brown, 2003).
Recent evidence suggests that there are numerous factors and individual
differences influencing the probability of being distracted (Wallace, 2003), with in-
vehicle distraction and external-to-vehicle distraction being the distinction in the
literature. With the emergence of more complex vehicle systems and electronic
devices available to drivers, overload of drivers is estimated to increase as new
technologies proliferate the market. Recent concerns about safety implications of
technology based distractions, centre on the magnitude and nature of demands these
devices place on drivers (Ranney, 2000, Min, 1998).
Distraction is frequently cited as another category of unsafe actions at level
crossing, with one U.S.A. study (National Transportation Safety Board., 1998) citing
motorist distraction as a probable cause in approximately 20% of collisions reviewed.
The types of distractions that have been documented in level crossing collisions
include the use of mobile phones; internal cognitive processes (e.g. day dreaming or
worrying); conversation with passengers; external distraction by objects or other
vehicles; and using in-vehicle devices (e.g. radio/CD player) (Caird, 2002). With the
advent of Intelligent Transport Systems in cars, level crossing collisions involving
distraction of motorists may become more prevalent. Currently, the Australian
Transport Safety Bureau records distractions as unintended motorist behaviour in the
data, and such behaviours are more common in level crossing collisions than in other
fatal road crashes. However, this data does not specify the incidence of distraction as
a major contributing factor to level crossing collisions in Australia nor the types of
distractions. According to Britain’s Health and Safety Executive (HSE) distraction at
level crossings may result in motorists failing to abide by warning signals and
crossing rules (Health and Safety Executive, 2005). The HSE (2005) suggests that
distraction issues may include:
• Noise: noisy surroundings may mean that users cannot hear oncoming trains;
• Groups: level crossing users in groups may behave differently to individuals,
for example when trying to keep together; and
• Visual distractions: visual distractions, such as shops and advertisements, may
draw the user’s attention away from the information and warnings at the
crossing.
91
2.8.9 Attentional blindness
According to Simons (2000) “although we might intuitively believe that
unusual, unexpected and salient objects will capture attention, leading to awareness,
they often do not” (p150). Attentional blindness is a major side-effect of top-down
processing (consciously directed attention in the pursuit of predetermined goals or
tasks), and in interactive tasks (such as driving), portions of the scene may go
unnoticed (Cater et al., 2002, Mack and Rock, 1998). This phenomenon denotes the
failure to see highly visible objects that a person may be looking at directly, when
attention is elsewhere (Mack, 2003, Memmert, 2006). Although typically a visual
phenomenon, there can also be an auditory and tactile component. The phenomenon
of ‘attentional blindness’ has been the topic of debate in level crossing safety in recent
years, as it has serious life-and-death consequences. It points itself to the serious
dangers of inattention whilst driving at level crossings, although it is difficult to
accurately determine the extent to which it may play a part in fatalities at level
crossings.
2.8.10 Hypovigilance and fatigue
Hypovigilance and fatigue have long been regarded as probable contributing
factors in many road crashes. Although vigilance and/or fatigue have not been
identified as major contributing factors in level crossing collision data per se, these
factors may have contributed to some collisions.
Hypovigilance has been defined as “a naturally occurring state in which some
individuals need more sensory stimulation to stay awake” (Dahl, 1996, p44).
Vigilance has been defined as having two broad conceptions. The first is related to
the physiological processes underlying alertness or wakefulness; while the second is
related to information processing and sustained attention (Thiffault, 2003). Fatigue
on the other hand is a general term used in the literature, which relates to both
physiological and psychological processes. According to Lyznicki et al (1998) fatigue
reflects a decreased capacity to perform, along with the subjective states which are
associated with decreased performance. Long hours spent driving (time-on-task) has
been found to produce fatigue and a deterioration of driving performance. A study by
Summala and Mikkola (1994) found that 60% of fatal sleep-related road crashes in
Finland occurred within the first hour of driving. The circadian rhythm (time-of-day)
92
effect has also been found to play a major role in fatigue (Pack et al., 1995).
According to Thiffault and Bergeron (2003):
The characteristics of road geometry and roadside environment, including other
factors that define the driving task, can have an impact on driving performance
by affecting arousal, alertness and information processing. An under-
demanding monotonous road environment with low traffic density can produce
fluctuations of arousal that decrease alertness and vigilance.
(p382)
As such, through the habituation process, low variation may tend to lead to
decreases in arousal, which may induce driving fatigue (Thiffault, 2003).
Countermeasures to rupture monotony have long been a topic of debate, with
the equilibrium between distraction and monotony of the foremost importance.
However, interventions discussed in the literature have been described as positive
distracters as they attempt to direct the motorist’s attention away from internal
thoughts and back to the external environment by providing a rupture in monotony.
There have been numerous suggestions as to how to rupture monotony. Brown (1991)
proposes that it could be beneficial to integrate novelty and variety into the motorist’s
task and environment, while Nelson (1997) suggests that partial perceptual
restrictions render the landscape more interesting and lessens boredom and monotony.
However, such suggestions have yet to be systematically evaluated in terms of their
effect on motorist fatigue and vigilance.
With the majority of studies indicating that fatigue is more frequently related to
monotonous road environments (such as driving on highways or on rural roads), it is
important not to rule out fatigue and hypovigilance as probable contributing factors in
passive level crossing collisions. Although rumble strips have been recommended by
Australian experts to alert drivers before they approach a level crossing, such trials are
only in the early stages of development. Visual stimuli prior to the placement of
rumble strips may be another option for countermeasure future development.
2.8.11 Speeding
Speeding has not been identified as a major contributing factor in level crossing
collisions, although may be one of several factors or influences in the chain of events
93
commonly leading to a collision. Speeding has been found to increase the incidence
and severity of road crashes (Elvik, 2004b, Australian Transport Council, 2000,
Buzeman, 1998, Johnston, 2004, Richter, 2004), with the severity of injuries resulting
from a road crash also being directly related to the pre-crash speed of the vehicle
(Frith, 2003).
Frith, Strachan and Patterson (2005) have summarised the relationship between
speed and road crash risk from a review of the literature.
• Between 1987 and 1988, 40 states in the United States of America raised the
speed limit on interstate highways from 55mph (88 km/h) to 65 miles/hour
(104 km/h). This resulted in an increase in average car speeds of about 3
miles/hour (5 km/h). Over the same period there was an increase in deaths on
these roads of between 20% and 25% (Transportation Research Board, 1998);
• During the 1973 fuel crisis, the New Zealand Government reduced rural speed
limits from 55 miles/hour (88 km/h) to 50 miles/hour (80 km/h), leading to an
8-10 km/h reduction in average rural speeds. The drop in speed led to a
significant drop in injuries, as compared with urban roads which were
unaffected by the speed-limit change. On main inter-city roads the number of
deaths dropped by 37%, serious injuries decreased by 24% and minor injuries
decreased by 22%. The corresponding reductions for urban areas were 15%,
9% and 4% (Frith and Toomath, 1982);
• In Australia the speed limit on Melbourne’s rural and outer freeway network
was increased from 100 km/h to 110 km/h in 1987 and then changed back to
100 km/h in 1989. Compared to a control area where the speed limit remained
the same, the injury crash rate per kilometre traveled increased by 25% when
the speed limit increased, and decreased by 19% when the speed limit
decreased (Sliogeris, 1992);
• A review of the studies on speed-limit changes from several countries (South
Africa, Belgium, Finland, France, Great Britain, Germany, U.S.A., and New
Zealand) where a speed limit was reduced or a new limit was introduced found
a reduction in road crashes ranging from 8% to 40% (Fieldwick, 1993) ; and
• Patterson et al. examined the effect of changes in speed limits on deaths on
rural interstates in the U.S.A. Road crash deaths in the groups of states that
raised their speed limits to 75 miles/hour and 70 miles/hour rose by 38% and
35%, respectively, relative to fatality levels in the states that did not change
94
their speed limits(Patterson et al., 2002).
(Frith et al., 2005, p24)
Fergusson et al. (2003) suggest that travelling above the speed limit is part of a
spectrum of risk taking behaviours that is associated with an elevation in crash risk.
Additionally, the concept of speeding forms part of violations of road rules, with
speeding being a deliberate deviation from the road rules (Kontogiannis, 2002,
Mesken, 2002, Parker, 1995b). However, the culture of speeding is virtually
perceived by most as ‘normal’, with many motorists believing that they can ‘safely
speed’ (Blincoe et al., 2006). Rothengatter (1991) proposes that many motorists
regard speeding as one of the least serious traffic offences, while Holland and Conner
(1996) suggest that speeding is seen as being socially acceptable and that motorists
believe that there is little chance of either being apprehended by police or causing a
road crash.
Police enforcement is one of the most frequently used tools in reducing speed
limit violations. The aim of enforcement is to deter motorists from speeding by
increasing one of the disadvantages of speeding – the perceived likelihood of being
caught (Frith et al., 2005). According to Goldenbeld and van Schagen (2005):
Given the fact that in many countries traffic law enforcement forms a central
part of the road safety program, there is still a considerable limitation in the
extent of the scientific knowledge about the most (cost-)efficient ways of
enforcing traffic violations in general and speed violations in particular.
(p1136)
While there is a large body of evidence that suggests that speed cameras can
reduce motorists’ speed as well as reducing frequency of collisions and severity of
trauma, not all motorists respond to speed cameras in the same way (Blincoe et al.,
2006). Corbett and Simon (1999) found in their study, that motorists react in a
complex way to speed related issues and speed camera enforcement. Their study
confirms findings from Corbett (1995) and Lex Service’s (1997) findings that
motorists can be categorised into specific groups (depending on their speed and
camera-related driving behaviour). The categories proposed include:
• Conformers – those motorists who always or nearly always comply with speed
limits. This group is most in favour of cameras;
95
• Deterred – those motorists who reduce their speeds since roadside cameras
have been in place. Cameras have made this group change their behaviour;
• Manipulators – those motorists who slow down for the camera then speed
outside the camera zone. Difficult group to target with anti-speed messages;
and
• Defiers – those motorists who exceed speed limits and have not reduced their
speed since speed camera introduction. Difficult group to target with anti-
speed messages.
(Corbett and Simon, 1999)
At level crossings, speed has not been identified as a major contributing factor
in level crossing collisions (albeit limited detailed data exists for level crossing
collisions in Australia). Nonetheless, speed may be one of several factors or
influences in the chain of events commonly leading to a level crossing collision.
2.8.12 Alcohol and drugs
Alcohol and drugs have been identified as contributing to 9% of fatal level
crossing collisions, compared with 30% in other fatal road crashes (Australian
Transport Safety Bureau, 2003). Although there has been a general reduction in
alcohol-related road crash fatalities and serious injuries during the past two decades,
drink driving remains a major public health problem for motorised countries
throughout the world (Sweedler, 1997). Drink driving is major factor in fatal road
crashes, with a significant percentage of fatal road crashes involving a driver with an
illegal blood alcohol content (Federal Office of Road Safety, 1997). Approximately
one-third of all fatal road crashes where the driver has been tested for alcohol
consumption are associated with blood alcohol concentrations (BAC) above the legal
limit (Ferguson et al., 1999), with approximately 70% of those drivers having a BAC
more than three times the legal limit (Single, 1997). Major contributing factors to the
decrease in road deaths has come about by improvements to roads and vehicles,
enactment of road safety legislation, intensive public education, and enhanced police
enforcement aided by improved enforcement technology (Federal Office of Road
Safety, 1997).
96
Over the past decade, there has been a substantial increase in driving fatalities
involving drug use. Drugs other than alcohol have been found to be associated with
approximately 23% of heavy vehicle driver fatalities (Degenhardt, 2006). According
to Drummer et al. (2004) in 2001, 16.5% of driver fatalities had used cannabis (THC)
or stimulant/amphetamine type drugs, whilst in 2002 this percentage had risen to
20.4%. International studies indicate that 5-30% of motorists involved in road crashes
were affected by drugs (Del Rio, 2002, Athanaselis, 1999, Longo, 2000, Sjogren,
1997). Evaluating the influence of drug use on driving is fraught with difficulty, with
research being restricted in several ways:
• Most subjects are young healthy males, who are not regular drug users;
• Drug concentrations used are low to moderate;
• Sample size is usually small;
• The validity of the tests used with regard to road safety is not clear; and
• The test situation in the laboratory results in an increased effort by the subjects
to compensate for drug effects.
(Krueger and Vollrath, 2000)
However, a recent case-control study by Drummer et al (2004) found that
cannabis, amphetamines and combinations of psychoactive drugs significantly
increase a motorist’s risk of a serious road crash. Although there is significantly more
research being conducted in this area, it is likely that little gains are to be made
specifically in the prevention of level crossing collisions with the development of
alcohol and drug countermeasures.
2.8.13 Gender
The relevance of gender to road safety has long been recognised and it has been
the contribution of male drivers to fatal and serious crashes which has, to date,
attracted the most attention.
(Dobson, 1998, p11)
Male drivers have historically been over-represented in road crash fatalities and
are more likely to be killed than female drivers for every kilometre traveled (Dobson,
1998). Statistics indicate that while male drivers may be more at risk of being
97
involved in a fatal road crash, female drivers are more at risk of sustaining serious
injury in a road crash (Federal Office of Road Safety, 1996). A project exploring
factors associated with risk of road crashes among young women (19-24 years) and
middle-aged women (46-51 years), found that “crash involvement is related to several
factors including feeling stressed and rushed, low life satisfaction, usual alcohol
consumption…and being born in a non English speaking country” (Dobson, 1998,
p9). Dobson et al (1998) suggests that errors, lapses and violations are related to both
the incidence of crashes and responsibility for those crashes in both young and
middle-aged women drivers.
2.9 RESEARCH QUESTIONS AND OBJECTIVES
This literature review highlights the fact that there is little known about the
exact causes of collisions at level crossings or about the effectiveness of reducing
collisions at crossings. The first study in the research program will examine the
following important research questions:
• What types of motorists are at risk of a collision?;
• What are the behaviours that motorists exhibit that increase their chances of a
collision?;
• How frequently do incidents and collisions occur?; and
• How frequently are incidents recorded by train drivers?
The overall objective of the second study is to provide information and
understanding about the knowledge, attitudes, norms, beliefs and risk perceptions of
motorists in relation to driving at level crossings. Specific research questions for
Study Two included:
• Do motorists perceive that they are at risk of being involved in a level crossing
collision?;
• What level of knowledge of the road rules exists?;
• What source and medium is believed to be the most appropriate for road
users?; and
• What differences exist between train drivers and experts opinions in terms of
high risk behaviours at level crossings?
98
Study Three involved three parts. The aim of Part One of this study was to
develop targeted interventions specific to each of the three road user groups in
accordance with Fishbein’s theoretical model (Integrated Model of Behaviour
Change). Part Two involved the investigation of the present context of unsafe driving
behaviour at level crossings. This included examining knowledge, attitudes, beliefs,
perceptions, self-reported and intended behaviour, and environmental constraints of
each road user group. This second part also involved the examination of the present
context of motorist behaviour at level crossings using key constructs from Fishbein’s
Integrated Model of Behaviour Change. Part Three involved trialing a pilot road
safety radio advertisement using an intervention and control methodology. This part
investigated the changes in pre and post-test constructs such as intentions, attitudes,
norms, self-efficacy or perceived behaviour control, perceived risks, and perceived
environment constraints whilst driving at level crossings.
2.10 SUMMARY
As can been seen, there is an extensive body of published literature related to
engineering measures for improving safety at level crossings. However, there is
limited published literature in the areas of human factors, education or enforcement
for this complex driving environment. This literature review has attempted to provide
a greater understanding of the possible critical factors involved in vehicle-train
collisions. Although vehicle-train collisions are rare events, and therefore it is
difficult to obtain valid information about preceding behaviour and the frequency of
near-misses, unintended road user error has been blamed by transport authorities as
playing a significant role in such collisions. However, anecdotal evidence also
suggests that risk taking at level crossings makes up a large proportion of near-miss
incidents.
It must be noted though that as long as level crossings present inherent dangers
to motorists (and as long as they exist), rail authorities will remain at risk of being
found liable (either in whole or in part) for fatal collisions (Stephen, 2002). With
more than 9400 level crossings throughout Australia, it is a difficult task (some would
say impossible) for rail authorities to provide engineering systems that will
completely protect all motorists from involvement in a collision. With this in mind,
99
educational interventions for improving motorist behaviour at level crossings may be
the answer that rail authorities are searching for. However to date, little rigorous
research has been conducted that has either explored the determinants of behaviour
and associated key constructs or measured changes in behavioural constructs or
statistically significant interactions between such constructs. Additionally, no
research has been guided by theory in exploring motorist behaviour at level crossings.
As a consequence, little is known about the role that educational interventions play in
changing intention of motorists. Without such research, it is difficult for transport
authorities to justify expenditure on campaigns targeting behaviour.
Although there are a number of models/theories that may be useful in guiding
this research on level crossing motorist behaviour (see Appendix 1), the ‘Integrated
Model of Behavioural Change’ (Fishbein, 2000) is the model chosen to assist in the
exploration of the entire program of research. Justification for the application of this
model to this program of research is provided. Although this model assists in all
aspects of this research program, it is particularly relevant to Study Three. It must be
noted that this model does not directly assess the predictive utility of Fishbein’s
model, but rather utilises it to assist in exploration of level crossing motorist
behaviour.
100
CHAPTER THREE: THE IMPORTANCE OF THEORY FOR IMPROVING MOTORIST BEHAVIOUR AT LEVEL CROSSINGS
3.1 Introduction …………………………………………………………. 101
3.2 The role of theory in road safety evaluation studies ………………... 101
3.3 The intention-behaviour relation ……………………………………. 102
3.4 Integrated model of behaviour change ……………………………… 103
3.5 Summary ……………………………………………………………. 106
101
3.1 INTRODUCTION
Explaining human behaviour in all its complexity is a difficult task which can
be approached at many levels (Ajzen, 1991). The motivation underlying road user
behaviour determines to a large extent how successful behaviour change strategies
may be. Although there are a number of models/theories that may be useful in
guiding this research program on level crossing motorist behaviour (as reviewed in
this chapter), the ‘Integrated Model of Behavioural Change’ (Fishbein, 2000) includes
two important constructs in driving: skills (or abilities) and environmental factors.
Fishbein (2003) suggests that the model recognises the lack of skills (or abilities)
and/or environmental constraints may prevent a person from acting on their
intentions, in light of the fact that intention is viewed as the primary determinant of
behaviour. While the majority of behaviour change theories are limited by a range of
conceptual and contextual factors (Parker, 2004), the IM was used to assist this
research program as it appeared to be the most applicable model to examine level
crossing safety.
While the IM proposed by Fishbein assists in the exploration of the entire
program of research, it is particularly relevant to Study Three. As no known research
has been conducted that utilises any psychosocial model to explain or predict level
crossing behavior within different road user groups, this research program used this
model as an exploratory tool rather than a tool to assess the model’s capacity in
explaining such behaviour. An extensive review of other models/theories and their
application to level crossing safety is provided in Appendix 1.
3.2 THE ROLE OF THEORY IN ROAD SAFETY STUDIES
In road safety research, some researchers argue that the role of theory has
historically been overlooked with many research programs having little theoretical
basis to guide the design of studies and the interpretation of study findings (Elvik,
2004a). Elvik (2004a) proposes that this deficiency in a strong theoretical basis means
that few results of road safety evaluation studies can be ruled out on theoretical
grounds. According to Risser and Nickel (2004), one of the major problems that has
contributed to this lack of a strong theoretical basis is the deficiency of meta-theory
that keeps the unstable and context-dependent theories under better control.
102
Additionally, Risser and Nickel (2004) suggest that “there is no, and there never will
be any, valid unified theory of human behaviour. Therefore, there will never be a
convincing unified theory of driver behaviour” (p52). More recently though, many
attempts have been made to develop theories designed to predict or explain the
findings of road safety evaluation research (Elvik, 2004a). These many theories have
been developed and used extensively to predict health behaviours which interpret a
person’s intention to act as the most immediate and important predictor of subsequent
action (Sheeran, 2004).
3.3 THE INTENTION-BEHAVIOUR RELATION
Understanding the intention-behaviour relation is fundamental as it offers a
method of organising and understanding the large number of influences within the
area of road safety (Gordon and Hunt, 1998). Intentions have been defined by
Triandis (1980) as the instructions that people give themselves to perform a particular
behaviour or to achieve certain goals. Such intentions are the culmination of the
decision-making process and according to Ajzen (1991), Gollwitzer (1990) and Webb
and Sheeran (2004) they signal the end of the reflection about a certain behaviour and
encapsulate the standard of performance that one has set for themselves, one’s
commitment to that performance, and the amount of time and effort that will be
expended during action. The construct of intention as being central to theories of
health behaviour also requires the understanding as to how well intentions actually
predict behaviour.
How well intentions predict behaviour and behaviour change has been the
subject of great debate by researchers and psychologists alike. Although many health
behaviour studies have found that intentions predict corresponding behaviour quite
well, there has long been a concern by researchers that there are observed
discrepancies between behaviour and intention (Sheppard et al., 1988, Campbell,
1963, Blumer, 1955). According to Gillholm et al. (2000) studies of the intention-
behaviour relation have primarily been concerned with the intention to perform single
behaviours, however in the real-world people commonly face a choice between a
multitude of behaviours.
103
3.4 INTEGRATED MODEL OF BEHAVIOUR CHANGE
The more one knows about the factors that underlie the performance or
nonperformance of any given behaviour, the more likely it is that one can
design successful interventions to change or reinforce that behaviour.
(Fishbein, 2003)
3.4.1 Development of the model
The ‘Integrated Model of Behaviour Change’ (IM) developed by Fishbein
(Fishbein, 2000), is a framework for “identifying the factors underlying the
performance or nonperformance of any behaviour clearly defined in terms of action,
target, circumstance and time” (2003, p3). Fishbein, a leading behavioural scientist,
suggests that although there may be an infinite number of variables that may directly
or indirectly influence the performance or nonperformance of a given behaviour, only
a limited number of variables are needed to be considered when predicting,
understanding, changing or reinforcing a given behaviour (Fishbein, 2003). Four
major theories of behaviour and behaviour change represent this limited set of
variables: (1) the health belief model, (2) social cognitive theory, (3) theory of
reasoned action, and (4) theory of planned behaviour (Fishbein, 2003). Based on
these four theories, Fishbein suggests that there are five key variables: (1) intentions,
(2) attitudes, (3) norms, (4) self-efficacy or perceived behavioural control, and (5)
perceived risk (Fishbein, 2003).
Also included in this model are skills (or abilities) and/or environmental
constraints. Fishbein (2003) suggests that the model recognises the lack of skills (or
abilities) and/or environmental constraints may prevent a person from acting on their
intentions, in light of the fact that intention is viewed as the primary determinant of
behaviour. He proposes that intentions alone are not the singular determinant of
behaviour, and different factors may satisfy or increase the intention/behaviour
relationship (Fishbein et al., 2003). In short, Fishbein proposes that intentions are
viewed as a function of the attitude towards performing the behaviour, perceived
norms (governing the performance of that behaviour) and self-efficacy (the person’s
perception of their ability to perform that behaviour when barriers to performance are
104
present). Each of these three main determinants is measured both directly and
indirectly.
According to Middlestadt (1996), a standard elicitation procedure is required to
develop indirect belief-based measures. This procedure identifies the significant social
referent groups that form the basis of normative orientation toward the behaviour, the
relevant barriers towards performance of the behaviour, and the underlying (indirect)
sets of anticipated outcomes that form the basis of attitudes toward the behaviour
(Middlestadt, 1996, Fishbein et al., 2003). Therefore, this model proposed by Fishbein
(2003) assumes that these fundamental beliefs determine the overall (direct) measures
of attitude, norm and self-efficacy. Also included in this model are distal variables
(such as demographics, personality characteristics) as they propose that they are
relevant predictors of intentions and behaviour, but are assumed only to affect
intentions (indirectly) through their influence on underlying beliefs (Fishbein et al.,
2003). The figure below illustrates the model proposed by Fishbein (2000).
Figure 8: Integrated Model of Behaviour Change
105
3.4.2 Criticisms of the model
Although the IM has not been used extensively in the literature, probably due to
its more recent development, there has been some validation of its worthiness in
predicting behaviour (Sheppard, 1988, Van den Putte, 1991). However, Fishbein
himself suggests that:
…a striking feature of the model is the complexity of the causal model of
intention formation compared with the apparent simplicity of the intention-
behaviour relationship. In fact, the IM does not actually address the ‘causes’ of
performing the behaviour except insofar as it defines intention as the primary
cause while recognizing the lack of ability and/or environmental constraints
may act negatively to prevent one from acting on one’s intentions.
(2003, p4)
Fishbein (2001) in his article examining utilising intervention theory to model
factors influencing behavior change, proposes that this does not mean that the
intention-behaviour correlation is low. He suggests that there are many studies that
show a substantial correlation whilst measuring both intentions and behaviour
(Fishbein et al., 2001), although he notes that such considerable correlations only
address the issue of statistical conclusion validity (Cook and Campbell, 1979), whilst
he suggest that the IM is a causal model of intention formation and ensuing
behaviour. From this however, he proposes that the model is a “complex causal model
of intentions but essentially a correlational model for behaviour” (Fishbein et al.,
2003, p5). Fishbein goes on to emphasise that this conception suggests an exploration
of the causal role of attitudes, norms and self-efficacy in performance of the target
behaviour, in contrast to forming positive intentions toward the behaviour (2003).
Other authors have suggested that the causal link between intentions and behaviour
could be intensified by other variables in the model. Some authors argue that
attitudes may have a direct influence on behaviour (Bentler, 1981, Kleinke, 1984,
Kraus, 1995), while Ajzen (1991) and Bandura (1986) argue that self-efficacy can
also affect behaviour directly (Fishbein et al., 2003). While the majority of behaviour
change theories are limited by a range of conceptual and contextual factors (Parker,
2004), this integrated model of behaviour change appears to be the most applicable
model to examine variables with regards to level crossing safety.
106
3.4.3 Application to level crossing behaviour
This model developed by Fishbein, has historically been utilised for research
examining HIV and STD prevention interventions. No known research has applied
this model to road safety. Therefore, this research is a ‘world-first’ in applying this
model to motorist behaviour at level crossings.
3.5 SUMMARY
The aim of this chapter was to present the merits and limitations of Fishbein’s
integrated model of behaviour change (IM) for exploring motorist behaviour at level
crossings and factors contributing to such behaviours. While the IM is relatively new
to the area of road safety, this model appears to provide a comprehensive framework
for understanding the variety of factors than can either encourage or discourage
unsafe driving behaviour at level crossings, than do other theories or models (as
presented in Appendix 1). Since behaviour change models have not been used in
exploring level crossing motorist behaviour, there is no doubt that this model
embodies an important perspective from which to examine level crossing driving
behaviour.
As stated previously, while the IM proposed by Fishbein assists in the
exploration of the entire program of research, it is particularly relevant to Study Three
for intervention and questionnaire development. As no known research has been
conducted that utilises any psychosocial model to explain or predict level crossing
behavior within different road user groups, this model was used as an exploratory tool
rather than a tool to asses the model’s capacity in explaining such behaviour. The
capacity to explain motorist behaviour at level crossings is enhanced through the use
of this model as it includes two important constructs in driving: skills (or abilities) and
environmental factors (Fishbein, 2000). With level crossings being high risk
intersections (often in complex driving environments), the inclusion of these two
elements for explaining behaviour is paramount.
107
CHAPTER FOUR: HIGH RISK AND VULNERABLE ROAD USERS AT LEVEL CROSSINGS
4.1 Introduction …………………………………………………………… 108
4.2 Review of existing data ……………………………………………….. 108
4.3 Study aims and research questions ……………………………………. 109
4.4 Method ………………………………………………………………... 110
4.5 Results ………………………………………………………………… 119
4.6 Discussion …………………………………………………………….. 124
4.7 Summary ……………………………………………………………… 125
108
4.1 INTRODUCTION
This chapter documents the first study undertaken as part of this research
program. This first study used the method of ‘triangulation’ (i.e. combining research
methods to give a range of perspectives) whereby both qualitative and quantitative
research designs were utilised. Two types of data collection methods were utilised in
this study. The first method (modified Delphi technique) involved a method of
structured collection and processing of information gathered from road and rail safety
experts in Australia to gain an informed judgment. This was designed to establish the
behaviours and types of road users that may increase their risk of being involved in a
level crossing collision. A modified Delphi technique is a particularly useful tool
when accurate information about a specific problem is limited or unavailable. The
second method (focus groups) involved the investigation of train drivers’ experiences
of incidents and collisions at level crossings. This provided an insight into the
incidents at level crossings to ascertain which road user groups are most at risk and
how they can be effectively targeted by educational countermeasures. Focus groups
with train drivers was seen as an important tool as it offers the opportunity to gain
valuable insights into participant’s perceptions and experiences, as well as providing
‘high density’ qualitative information (Nicolson and Anderson, 2003).
Much of what is currently known about level crossing collisions is derived from
coroner’s findings and statistics. This study is designed to extend this knowledge by
undertaking a more thorough examination of unsafe behaviours and the road user
groups at risk.
While the principal aim of this study is to address the specific research
questions identified below, it will provide a foundation for Study Two and Study
Three. These later studies will investigate the effectiveness of an educational
countermeasure for three road safety groups identified to be at risk by the current
study and examine the present context of motorist behaviour at level crossings.
4.2 REVIEW OF EXISTING DATA
Currently, exact statistics of level crossing fatalities in Australia are difficult to
determine. The most significant reason is that most jurisdictions in Australia use
different methods of categorising and recording such collisions. As a consequence of
109
these differing methods, there is a lack of definitive evidence available relating to the
extent and nature of level crossing collisions. Due to the large number of incidents at
level crossings that do not involve a fatality but have the potential to result in a
catastrophic event such as a train derailment, it is important to examine the
occurrences of incidents as well as collisions involving fatalities. To date, the
Australian Transport Safety Bureau is unaware of the exact number of level crossing
near miss rates in each jurisdiction (Australian Transport Safety Bureau, 2003),
however anecdotal evidence from train drivers suggests that near misses are far more
frequent than reported.
Due to this relatively small number of fatalities each year involving collisions
between trains and motor vehicles and limited data availability from the ATSB, it is
difficult to determine the major causes of collisions. However, the ATSB data does
indicate that older drivers (aged 60 years and over) are over-represented in level
crossings collisions compared to other fatal road crashes (Australian Transport Safety
Bureau, 2003). However, there are likely to be other road user groups that are also at
risk of being involved in a level crossing collision. With the number of heavy
vehicles on the road expected to increase dramatically over the next decade, the risk
of a heavy vehicle-train collision (that results in a train derailment) is also expected to
increase. Therefore, since there is very little useful data available to determine key
road user groups at risk of a vehicle-train collision, this study’s purpose was to
identify such road user groups through two data collection methods.
4.3 STUDY AIMS AND RESEARCH QUESTIONS
The purpose of this study was to identify road user groups at risk and key unsafe
motorist behaviours. As described above, there are a small number of fatalities each
year involving collisions between trains and motor vehicles and limited data
availability from the Australian Transport Safety Bureau (ATSB). Hence, it is
difficult to determine the major causes of level crossing collisions. Evidence from rail
authorities in all jurisdictions indicate that there are a small number of collisions but a
large number of ‘near-misses’.
Specific research questions for this study included:
• What types of motorists are at risk of a collision?;
110
• What are the behaviours that motorists exhibit that increase their chances of a
collision?;
• How frequently do incidents and collisions occur?; and
• How frequently are incidents recorded by train drivers?
4.4 METHOD
4.4.1 Overview
Two forms of data collection (quantitative and qualitative) were used to provide
methodological triangulation and ensure reliability of the data. The triangulation
method uses multiple methods and perspectives to collect and interpret data to seek
convergences of findings to enhance validity. This study consisted of two phases: (1)
consensus of an expert panel (modified Delphi technique) and (2) focus group
discussions with train drivers.
4.4.2 Delphi technique
4.4.2.1 Background information
When accurate information is unavailable, as in the field of level crossing safety
in Australia, the Delphi technique is a particularly useful tool. The Delphi technique
is one method that has been used to determine priorities, set goals and establish future
directions in a specific area (Linstone, 1975). Additionally, it is useful for non-
interacting groups (whose members are geographically distant) and for ‘hot topics’
facing the community (level crossing safety is currently one ‘hot topic’ in both the rail
industry and community).
Consensus methods are one technique in dealing with conflicting scientific
evidence (Jones, 1995). According to Jones and Hunter (1995), such methods allow a
wider range of study types to be considered than is customary to statistical reviews. In
addition, these methods allow a superior function for the qualitative assessment of
evidence. One of the best known consensus methods is the Delphi technique. The
Delphi technique takes its name from the Delphic oracle's skills of interpretation and
foresight (Jones, 1995). The Delphi technique is particularly useful when accurate
information is unavailable. It consists of a series of sequential questionnaires or
111
‘rounds’, interspersed by controlled feedback that seeks to gain the most reliable
consensus of opinion of a group of experts (Linstone, 1975). The first round
questionnaire is qualitative and its role is to identify issues to be addressed in later
rounds. The second and subsequent rounds are more specific, with the questionnaire
seeking quantification of earlier findings, usually through rating or ranking
techniques. Each participant is provided with a summary of opinions before
answering the next questionnaire. The survey is the most common technique of
Delphi application. The Delphi survey has three special features: 1) anonymity of
participants, 2) iteration and controlled feedback between rounds, and 3) statistical
summary of group response (Linstone, 1975). Adler and Ziglio (1996) give three
considerations that are important for Delphi applications:
• The problem does not lend itself to precise analytical techniques but can
benefit from subjective judgments on a collective basis;
• The problem at hand has no monitored history nor adequate information on its
present and future development; or
• Addressing the problem requires the exploration and assessment of numerous
issues connected with various policy options.
Typically, the Delphi technique commences with an open-ended questionnaire
that is given to a panel of selected experts to solicit specific information about a
particular subject or content area (Custer et al., 1999). In subsequent rounds of this
procedure, participants rate the relative importance of individual items as well as
making changes to the phrasing or substance of the items. Through a series of rounds
(typically three) this process is designed to yield consensus of the subject (Custer et
al., 1999). A modified Delphi technique is similar to the usual Delphi technique in
terms of both procedure and intent. However, the major modification consists of
commencing the process with a set of meticulously selected items. Review of the
literature and interviews with experts in the area, are the two main sources from
which pre-selected items are drawn (Custer et al., 1999). Custer et al. (1999) suggests
that the primary advantages of this modified Delphi technique is that it provides solid
grounding in previously developed work, and it improves the initial round response
rate.
112
The purpose of these survey questionnaires was to identify what experts believe
are the key road user groups and target behaviours that are risk factors for a vehicle-
train collision at a level crossing, as no national dataset is available in Australia that
accurately provides detailed demographic or behavioural characteristics of drivers
involved in collisions. Studies by the ATSB provide only a description of some of the
circumstances that appear to be prevalent in fatal collisions at level crossings but in no
way claims to be definitive of the nature and causes of level crossing collisions
(Australian Transport Safety Bureau, 2002b). The ATSB (2002b) states that the
sample sizes are too small to enable in-depth analysis.
In road safety, the Delphi technique has been used to assess expert opinion on
countermeasures against unlicensed driving (Knox, 2002) as well as speeding (Silock,
1999). Based on the above criteria, characteristics of the Delphi technique were
considered particularly suitable for the consideration of identifying road user groups
at risk in level crossing collisions. As no data is currently available in Australia that
accurately provides demographic or behavioural characteristics of motorists involved
in vehicle-train collisions, the Delphi technique was one tool that may assist in
informing the direction of Study Two and Study Three.
4.4.2.2 Sample and procedure
A list of experts in level crossing safety in Australia was compiled in
consultation with the Australasian Railway Association and state rail authorities. This
list included experts from the following agencies:
• Australian Rail Association (ARA)
• Australian Rail Track Corporation (ARTC)
• National and State level crossing safety steering groups (e.g. National Railway
Level Crossing Safety Strategy Implementation Group, QLD Level Crossing
Safety Steering Group)
• Departments of Planning and Infrastructure (DPI)
• Queensland Rail (QR)
• Queensland Transport (QT)
• Main Roads, WA
• Main Roads, Queensland
• VicRoads
• National Road Transport Commission (NRTC)
113
• Queensland University of Technology (QUT)
• Pacific National
• Monash University Accident Research Centre (MUARC)
• Australian Road Research Board (ARRB)
• State Coroner’s Offices
• N.S.W. Road and Traffic Authority (RTA)
• Police Departments
• Parliamentary TravelSafe Committees
• RACQ/RACV/NRMA
Survey questionnaires (with Information Sheets) were distributed to 48 experts
(both rail and road industry experts) throughout Australia. It was determined that rail
and road safety experts in level crossing safety from different regions of the country,
as well as industry experience and perspective would provide some breadth of insight
into level crossing safety issues. The rationale for using these two different industry
panelists was that although both groups came from different parts of the industry, they
represent unique perspectives of expert opinion. Both perspectives are important as
they will assist in informing development of educational interventions. An
explanation of the objectives and procedure of each Delphi round is detailed below:
• Round 1
Unstructured and open-ended questionnaire;
Experts can express their opinion on any issue they feel relevant;
and
Questionnaire contains a synopsis of the issues and is educative,
informative as well as being explorative.
• Round 2
Questionnaire developed from feedback in Round 1;
Quantitative statements that participants rated;
Answers from Round 1 fed back to participants through a summary
at the beginning of Round 2 questionnaire;
Self-rating of participants on their level of expertise; and
Final analysis and conclusions drawn.
114
Round One of this survey (48 surveys sent) was sent out by mail on the 28th
July 2004. A reminder letter was sent out to all experts on the 10th August 2004 to
encourage them to participate. Consent was implied if the questionnaire was
completed and returned. A total of 24 experts returned a completed questionnaire
(response rate of 50%). The Round Two survey was then developed with feedback
from the Round One survey and was sent on the 30th September 2004 to 45 experts (3
experts stated that they were unable to participate in the survey for unknown reasons).
A reminder letter was sent out to all experts on the 13th October to encourage them to
participate. Consent was implied if the questionnaire was completed and returned. A
total of 27 surveys were returned (response rate of 60%).
4.4.2.3 Assessment tools
As stated earlier, a modified Delphi technique was used for both rounds. The
first round included open-ended questions to solicit specific information about a
particular subject or content area. This first round questionnaire included carefully
selected items drawn from both the literature and interviews (informal) with experts in
the field. Round Two questionnaire participants rate the relative importance of
individual items as well as making changes to the phrasing or substance of the items.
Round One of the modified Delphi technique was developed to elicit broad
responses to questions about warning and protection systems; types of motorists at
risk of collision; unsafe motorist’s behaviours; and demographics. Opinions and
recommendations from Round One were provided to the panel and then displayed in
rating items in Round Two. The second questionnaire (Round Two) was quantitative,
and was developed to obtain ratings (using Likert scales) of the types of motorists at
risk and unsafe motorist behaviours. These Likert scales ranged from 1 = Very
Important to 5 = Not at All Important. Pilot testing of the two questionnaires was
conducted to help to identify ambiguities and improve feasibility of administration.
Both of these questionnaires were piloted on road and rail safety practitioners and
researchers.
4.4.2.4 Analysis
Opinions and recommendations from Round One were provided to the panel
and then displayed in ranking items in Round Two. This second round was used to
115
elicit opinions on more specific items than in the first round. Descriptives were
conducted using the open-ended answers in the Round One survey. Data from the
Likert scales from Round Two were entered into the Statistical Package for Social
Sciences (SPSS) and descriptive analyses were conducted (i.e. mean and standard
deviation etc.).
4.4.3 Focus groups
4.4.3.1 Background information
Focus groups with train drivers were seen as a critical element in examining the
prevalence of ‘near miss’ incidents. The purpose of this second phase is to explore
train drivers’ experiences of motorist behaviour at level crossings. As train drivers
potentially have a different perspective of the road user groups and key target
behaviours at risk of a vehicle-train collision from experts in the area, exploring train
driver’s experiences and opinions is seen as a valuable method of data collection. No
known research has been conducted in Australia that has explored the perceptions of
train drivers specifically in relation to level crossings. This phase focused on train
drivers’ experiences of:
• Motorist behaviour at level crossings;
• Vehicle types involved in level crossing incidents;
• “Near-misses” at level crossings;
• Perceptions on motorist behaviour at level crossings;
• Perceptions of current safety and safety actions at level crossings; and
• Impact of level crossing incidents.
Focus groups are one of the most favoured and accepted methods of collecting
qualitative data, as they offer the opportunity to gain valuable insights into
participant’s perceptions and experiences (Nicolson, 2003). Focus groups allow data
to be collected from a number of people in a relatively short period of time (Beyea,
2000) as well as providing ‘high density’ qualitative information (Nicolson, 2003).
Focus groups were selected as the most appropriate data collection too with train
drivers, as they offered the opportunity to explore and gather rich data on level
crossing safety as well as gaining insights into train drivers’ experiences. The group
116
experience with trains drivers is also advantageous as it facilitates discussion within a
supportive environment (Nicolson, 2003). Kitzinger (1995) states:
The idea behind the focus group method is that group processes can help people
to explore and clarify their view in ways that would be less easily accessible in
a one to one interview…When group dynamics work well the participants work
alongside the researcher, taking the research in new and often unexpected
directions.
(p299)
According to The Health Communication Unit at the University of Toronto
(2002), there are three main types of focus groups:
• Exploratory focus groups – used to increase understanding of an issue, to
generate hypotheses, in concept development and pilot testing;
• Phenomenological focus groups - seek to understand the experiences and
outlook of respondents (as consumers, potential consumers and/or opinion
leaders); and
• Clinical focus groups – used to examine unconscious mechanisms operating
within people that impact on their behavioural or predispositions to behaviour.
However, each of these types of focus groups has both advantages and
disadvantages. Advantages of focus groups include:
• The main advantage of focus group methodology for collecting information is
that it allows for in-depth discussion and probing on an issue of interest. You
can collect opinions of more than one person in one session and the interaction
between group participants can result in increased elaboration on a topic and
broader insight into understanding an issue;
• Focus groups provide a tremendous amount of information at a reasonable
cost. In most cases they are less costly than conducting 8–12 in-depth
interviews and cheaper than most quantitative data collection methods;
• Opinions of more people are obtained within a shorter time frame, compared
to in-depth interviews; and
• Clients can benefit by observing the group if a room with a one way mirror is
used.
(Kitzinger, 1995)
117
Whilst disadvantages include:
• Potential for participants to influence one another’s opinions;
• Focus groups do not provide quantifiable information about a population;
• The number of questions which can be asked is limited as the response time
for each question is increased by the number of participants;
• While focus groups provide the researcher with in-depth responses to their
questions, this type of data is more difficult to analyse than quantitative data;
• The quality of the information collected is dependent on the skills of the
moderator;
• For some populations and topics focus groups are not effective because the
social context influences the responses more than what the researcher would
want; and
• Focus groups can be difficult to conduct with populations which have hearing,
cognitive or communicative impairments.
(Kitzinger, 1995)
4.4.3.2 Sample and procedure
The subject pool for focus group discussions with train drivers was identified by
Queensland Rail (QR) through the QR Research Manager. Train drivers were
informed of the opportunity to participate in the focus groups by QR managers and
Union representatives. Information sheets and consent forms were provided to the
participants prior to commencement of the focus group. The focus groups took place
during October 2004, with one focus group being held in Rockhampton
(Regional/Freight services) and another focus group being held in Brisbane
(Citytrain). Seventeen train drivers from QR participated in the study: eight (8) from
the Brisbane metropolitan area and nine (9) from regional areas of Rockhampton and
Gladstone.
Each focus group discussion ran for approximately 90 minutes and moderated
by the same two researchers. Participants were informed that participation was purely
voluntary and that their responses would remain anonymous through de-identification
of any collected data. Information sheets were provided to the participants by the
Traincrew Managers prior to the focus group and posters advertising the sessions
were placed on noticeboards in the Traincrew staff areas. Consent forms were signed
118
by participants at prior to commencement of the discussion. Contact details of the
counseling service at Queensland Rail were provided to each of the participants on the
information sheet.
The first session was conducted in a provincial city with train drivers from
regional/freight service areas and the second with train drivers from the metropolitan
area. Both focus groups were conducted on Queensland Rail premises in a private
meeting room.
Each group had a full discussion of each of the items on the agenda and all
respondents were given sufficient opportunity to air their views. The researchers’
main objective was to facilitate discussion between participants. To allow external
validation, both group discussions were tape recorded and subsequently transcribed.
Written notes were also made during the discussion. Transcribed data from these
tapes will be de-identified to protect each participant’s identity. No theoretical testing
or development was necessary for these focus groups, as this study is exploratory and
the aim of the focus groups was to gather information on the frequency and
characteristics of level crossing incidents as experienced by the train drivers. Codes
were generated from the data collected from these two focus groups, with themes
being identified independently for each focus group and then merged together for
sorting of similar themes. During each session, an inductive approach was used,
which allowed the researchers to be flexible in exploring issues and themes as they
arose.
In consideration of the sensitive nature of involvement in accidents and
fatalities, participants were not asked directly if they had been involved in a vehicle-
train collision but were encouraged to discuss their experiences.
4.4.3.4 Assessment tools
An agenda for the focus group discussions was developed in accordance with
guidelines for conducting focus groups. This agenda included: welcome, review of
agenda, review of goal of the meeting, review of ground rules, introductions,
questions and answers, closing comments. Questions were developed with input from
experts in the area and were based on the themes identified from the survey
questionnaire (consensus of expert panel) results. The types of questions included:
nature, frequency and reporting of incidents at level crossings, types of vehicles
119
involved in incidents, types of behaviours observed at level crossings, the impact of
incidents, and current protection systems.
4.4.3.5 Analysis
Raw data collected from the focus group discussions was collected in a
relatively unstructured form (by use of tape recordings) and then transcribed. The data
was then analysed through a qualitative thematic analysis process.
4.4.4 Ethical clearance
Ethical clearance for data collection for Study One was gained from the
Queensland University of Technology Human Research Ethics Committee (QUT Ref.
No. 3550H).
4.5 RESULTS
4.5.1 Delphi technique
The response rate for Rounds One and Two were 50% (n=48) and 60% (n=45)
respectively. In Round Two, participants were asked how many years experience
they had in the field, with a mean of 22.8 years (S.D. 12.57) being observed. The
division between those participants who were working in rail, road, road and rail, and
enforcement was 51.9%, 25.9%, 7.4% and 14.8% respectively.
Themes from Round One included experts viewing the most common motorist
issues as: risk taking, inattention, disobeying road rules, complacency, motorist error,
low expectations of trains, and lack of education. The engineering and design issues
that experts believed contribute to collisions at level crossings included: sighting
distances, short stacking problems (particularly with trucks) and angles of approach.
One environmental issues was raised by many experts with regards to contributing to
level crossing collisions: the position of a crossing may restrict some motorists from
being able to see activated warning systems (such as flashing lights) when the sun is
shining brightly.
120
Many experts made comment on safety management issues, with some raising
the need for consistency of approach with understanding of the complexity of level
crossing safety. One expert commented:
“The interaction between a train and a vehicle or other road user is a complex
event. The vehicle, road user, or the infrastructures are the primary elements of
railway level crossings. These operate within a wider environment of
economics, community culture, scientific understanding, etc. All of these
contribute to potential conflicts.” - Participant Round 1.
The road user groups identified by the expert panel as being ‘at risk’ included:
general motorists, younger and older drivers, heavy vehicles, rural road users, buses,
fleet and local drivers. Apart from the general motorists group, there were four
groups that were identified by the majority of respondents as being ‘high risk’. These
included heavy vehicles, rural road user, older drivers and younger drivers. There
were four key areas that participants raised that they believed were major motorist
factors that contribute to collisions at level crossings. These included:
• Behaviour – some drivers try to beat trains or drive around boom gates;
• Training – drivers are not trained to deal with level crossings;
• Understanding – consequence and severity of accidents with trains; and
• Inattention – train never comes at this time so on auto pilot.
A range of behaviours and issues associated with these groups were identified
and utilised in the ranking processes of the Round 2 questionnaire.
The Round Two questionnaire was developed from the first round
questionnaire, with responses being re-introduced to the panel as factors for rating
(Likert scales). Results from this second questionnaire highlighted the importance of
specific items within the constructs of ‘major contributing factors’ and specific
behaviours and issues of each ‘road user group’. Within the construct of major
contributing factors, motorist behaviours that were found to be ‘very
important/important’ included:
• Low expectation of coming across a train while driving (89.3%);
• Not slowing down to scan for a train at passive crossings (85.7%);
• Not stopping at Stop signs at passive crossings (85.7%);
121
• Inattention by the motorist when driving (82.1%); and
• Lack of detection devices for breaches (no “red light” cameras at level
crossings) (75.0%).
For each of the road user groups a number of items that may contribute to
increased risk were presented for ranking. For the older driver group, there was a
high level of agreement that errors in judgment (e.g. misjudging time needed to cross
safely) were ‘very important/important’ risk factors (78.6%). For the younger driver
group, 92.8% ranked trying to beat the train across the crossing as ‘very
important/important’. Factors for the heavy vehicle group that were ranked as ‘very
important/important’ included: trying to beat the train across the crossing (75.0%)
and length of vehicle causing overhang on the crossing (67.9%). Factors particular to
rural road users that were ranked as ‘very important/important’ included: low
expectation of a train (89.3%); complacency due to familiarity (89.3%) and not
scanning for a train at give way signed crossings (85.8%).
4.5.2 Focus groups
All participants in the focus groups were male, reflecting gender profile of this
workforce. The mean years of industry experience for the metropolitan group was 24
years (range 1 to 34 years) and for the regional group was 23 years (range 5 months to
42 years). The majority of participants were train drivers (n=15), while two
participants were Train Guards. The majority of the drivers revealed they had
experienced an incident or fatality (including suicides) during their career with two
drivers reporting they had experienced more than five fatalities each. All participants
had experienced near misses with road users at level crossings.
Analysis revealed that there are strong differences between the experiences of
Regional/Freight participants and the Citytrain participants.
4.5.2.1 Citytrain
The metropolitan train drivers generally experienced motorist behaviour at
active crossings with flashing lights and boom gates while the regional train drivers
experienced behaviours at active crossings with boom gates, crossings with lights
only and passive crossings with stationary signs.
122
In the metropolitan train driver group, experiences of motorist behaviour at
level crossings included: motorists driving around boom gates, getting stuck under
boom gates, queuing over congested crossings and driving through the crossing after
the red lights commence flashing. The behaviour of motorists driving around boom
gates was noted to occur quite regularly with one participant commenting it happens
“everyday…they reverse back and then drive around them”. The majority of train
drivers believed that it was very common that motorists would drive through the
crossing when the lights are flashing both before and after the booms were activated
and some crossings were named as black spots where motorists repeatedly offend.
Vehicles protruding into the path of the train and motorists entering congested
crossings and then panicking and driving backwards into the boom gates were also
mentioned.
For the metropolitan group, trucks were also mentioned, with the issue of
getting stuck under the boom gates due to the overhang of the vehicle on the crossing.
Comments were made that in some cases the boom gates have closed on the rear
trailers of trucks without the truck driver even realising. Motorcyclists trying to beat
the train were mentioned as frequent high risk road user behaviour.
4.5.2.2 Regional/Freight
With regards to motorist behaviour, the regional group participants noted that
motorists not stopping or giving way at passively controlled crossings is continuing to
be a major concern and that behaviours differ with location. Interaction with
engineering devices was discussed and although it was agreed that generally people
stop for the barriers, high risk motorist behaviours at active crossings included
running the flashing lights and disobeying the boom barriers by driving around or
straight through them. Other high risk behaviours included motorists attempting to
beat the train across the crossing, speeding up at flashing lights, and general risk
taking by younger drivers in particular. Participants spoke of motorist behaviour that
they perceived were due to inattention such as, getting stuck under lowering boom
gates and not seeing the train and subsequently driving into its path. Motorists not
allowing enough time to cross in front of the train and hesitating or stop-starting at the
crossing were also noted to be at high risk. There was a general perception from these
123
behaviours that motorists are unable to judge the speed and distance of an
approaching train to determine a safe gap during which to cross.
Mention was made of risk taking in motorcyclists and younger drivers, however
the main cause of concern was heavy vehicles, as breaches at level crossings are
common, and the potential of a heavy vehicle-train collision is highly likely to
injure/kill the train driver and possibly derail the train. Heavy vehicles were noted to
be slower to get across the crossing due to vehicle length and acceleration. High risk
behaviour in this road user group included not stopping at passive crossings, trying to
beat the train across the crossing and going around or through boom gates.
There was agreement by the regional group that about 60% of ‘near misses’ are
due to poor visibility (such as angle of approach) with only 30-40% actually due to
the motorist. Perceptions on why motorists engage in risk taking behaviours at
crossings were a common theme throughout the discussion. These perceptions were
generally based on the train driver’s interpretation of the vehicle approach and
reactions of the motorist. In the cabin of a train, train drivers are often able to see the
vehicle on approach to the crossing and make predictive judgments about their
behaviour. Common thoughts included: “You see a car or truck… coming up the
road…and you think ’Is this bloke gonna stop or what?' ”. On seeing the motorists in
the vehicle, interpretations of frustration and impatience were made, such as:
You can actually see the look on the motorist’s face…when you’re on a train,
the boom gates come down and they are... ‘Aw the train’ they’re not real happy,
especially when they are going to work.
A thorough range of perceptions on why motorists enact in risk taking
behaviours at crossings were given by the participants including those factors that are
attributable to the environment rather than human factors. There was general
agreement in the regional group that motorists’ behaviour was influenced by the
motorists’ complacency, lack of knowledge and low perception of risk. Impatience
by the motorist was a commonly recognised theme with comments that, being in
hurry, time pressures, and anger and frustration with waiting for the train to pass,
being noted by both groups of participants.
For the regional train drivers, possible reasons for local motorist behaviours
included low expectations of coming across a train due to the infrequency of trains at
crossings (and subsequent complacency) and knowledge of the waiting times when
124
stopping for a train leading to beat the train behaviours. This perception of motorists
trying to beat the train, lights or boom gates to save time was a recurring theme with
both groups.
The lack of enforcement of the road rules at crossings emerged as a strong
theme in both the Citytrain and regional groups, with participants noting that
motorists know they won’t get caught. This was also suggested for the small amount
of motorists that engage in risk taking or thrill seeking behaviours. In the regional
group, motorist inattention and distraction were common issues identified, with the
concepts of motorists’ not seeing the train or not registering that they are at a crossing
mentioned. Motorist knowledge was also thought to be a contributing factor to
behaviour with comments that there is low knowledge of a train’s stopping distance,
public misconceptions on the ability of the train to stop and the poor understanding of
the meaning of warning lights. One train driver commented, “…people say that to
me, 'Why don’t we stop for them?' ”. Due to the high number of motorists that do not
leave enough time to cross safely, it was also thought that motorists did not have the
ability to judge the distance and speed of an approaching train to determine a safe gap
during which to cross. Perceptions on the behaviours of truck drivers included
possible influences of fatigue, time pressures, delays in waiting for long trains and
frustration, drivers not knowing the length of their vehicle, and misjudging the time
taken to cross safely.
For both the regional and metropolitan train driver groups, there appeared to
exist a culture that train drivers are ‘whingers’ if they report all near-miss incidents at
level crossings. Many train drivers in both groups indicated that the reporting
mechanism for such events (i.e. transmission over the radio to control) does not allow
any anonymity and thus they are reluctant to report the numerous incidents that they
encounter.
4.6 DISCUSSION
4.6.1 Study limitations
This first study used ‘methodological triangulation’ (i.e. combining research
approaches to give a range of perspectives) whereby both qualitative (focus groups)
and quantitative (modified Delphi technique) research designs were utilised (Barbour,
125
1999, Bryman, 1992). With the discipline of road safety research requiring
methodological strategies that will enhance efforts to conceptualise the multi-faceted
nature of motorist behaviour at level crossings, this application provided the
robustness required. However, like all research, this first study is not without its
limitations.
One of the foremost limitations of focus groups include the responses of each
participant potentially not being independent as a few dominant focus group members
can skew the session. However, with the experience and skills of the moderator, it is
anticipated that this limitation was reduced substantially. Additionally, analysing
results from these focus groups requires proficiency and a good understanding of
qualitative research. With the author having many years of experience in focus group
analysis in both health research and road safety research, it is anticipated that this
limitation was also reduced considerably.
With regards to limitations of the Delphi technique, it has been suggested that
success depends of the quality of the participants (Linstone, 1975, Masser and Foley,
1987). Identifying participants who are knowledgeable about level crossing safety
was a challenge. Although it was determined that rail and road safety experts in level
crossing safety from different regions of the country would provide different
perspectives to gain ensure a breadth of insight into level crossing safety issues, it is
difficult to guarantee that the choice of these experts was unsurpassed. However,
with the assistance of the Australasian Railway Association (ARA) in the compilation
of this list of experts, it is anticipated that the quality of these experts was of a high
standard.
4.7 SUMMARY
This study was designed to explore the first four research questions identified as
part of this program of research. In doing so, it identified road user groups at risk and
their key target behaviours in level crossing collisions. Results from methodological
triangulation research indicate that there are three main road user groups at risk: older
drivers (60+ years), younger drivers (17-24 years) and heavy vehicles. These results
extend the available evidence relating to contributing factors involved in vehicle-train
collisions. Additionally, they provide justification for targeting specific road user
126
groups in attempting to change behavioural constructs necessary for safe driving at
level crossings.
127
CHAPTER FIVE: PLANNING AND DEVELOPMENT OF INTERVENTIONS FOR EACH ROAD USER GROUP
5.1 Introduction …………………………………………………………… 128
5.2 Vulnerable road users at level crossings ……………………………… 129
5.3 Qualitative research with target groups ……………………………….. 148
5.4 Quantitative research with train drivers and experts ………………….. 173
5.5 Framework for intervention development …………………………….. 186
5.6 Summary ……………………………………………………………… 190
128
5.1 INTRODUCTION
This chapter documented Study Two, the formative research undertaken as part
of the planning, development and delivery of behavioural interventions for each of the
three road user groups. Additionally, it provided an overview of the road safety issues
specific to each of the three road user groups identified in Study One: younger
drivers, older drivers and heavy vehicle drivers.
Formative research uses social science methods to assess the beliefs,
perceptions, and behaviours of a specific group (Vastine et al., 2005). The resulting
data allow for the design and development of an intervention that is tailored to the
group’s requirements and preferences (Vastine et al., 2005). Additionally, formative
research demonstrates an interest in understanding target groups and can in theory
build trust, collaboration and cooperation, and ultimately acceptance of the project
(Vastine et al., 2005). This type of research forms the basis for developing effective
strategies, including communication channels, to try to influence behaviour change.
Two forms of formative research data collection (qualitative and quantitative)
were used to provide methodological triangulation and ensure reliability of the data.
As discussed in the previous chapter, this method uses multiple methods and
perspectives to collect and interpret data to seek convergences of findings to enhance
validity. Qualitative data collection involved formative research with members from
each of the three road user groups using semi-structured interviews and focus group
discussions from both urban and regional samples. Quantitative data collection
involved both train drivers in regional and urban settings, and experts in the field.
The same recruitment methods used in Study One for train drivers and experts in the
field were used in this second study. As both train drivers and experts assisted with
identifying the three road user groups at risk in Study One, they were also deemed to
be valuable in assisting with formative research in this second study. The overall
objective of this study was to obtain rich data on the key variables of attitudes, norms,
self-efficacy (perceived behavioural control), perceived risk, environmental
constraints and the skills/abilities of drivers and prioritise project resources for
intervention planning, development and delivery. Information obtained from both
forms of data collection (i.e. qualitative and quantitative) was critical in assisting,
guiding, and identifying priority areas for message and material development and
development of data collection measures (i.e. questionnaires).
129
Typically there are two key approaches to planning behavioural interventions.
One is ‘social marketing’, which is often used by transport agencies to plan mass
media campaigns, the other ‘intervention mapping’ is based on the importance of
planning behavioural interventions that are founded on theory and evidence. Both
approaches will be discussed in this chapter. Additionally, findings from the
formative research are discussed in detail.
5.2 HIGH RISK AND VULNERABLE ROAD USERS
5.2.1 Overview
Findings from Study One indicated that there are four main road user groups at
risk of being involved in a level crossing collision. These groups include younger
drivers, older drivers, heavy vehicle drivers and rural road users. As rural road users
comprise a number of distinct road user groups, including older, younger and heavy
vehicle drivers, rural road users are not examined per se. While not exhaustive, this
examination will provide an overview of the characteristics of the contributing factors
to general road crashes are described below, with application to level crossing safety.
Reference is also made to the literature (see Chapter Two – Literature Review) that
examines human factors contributing to road crashes (or level crossing collisions).
5.2.2 Younger drivers
The over-involvement of young drivers in road crashes is widely understood in
Australia to be a most serious and to date largely intractable road safety
problem. Young people aged 15-24 make up 15% of the population but account
for 31% of fatalities.
(Triggs, 1998)
5.2.2.1 Crash statistics and patterns of younger drivers
A study by Cavallo and Triggs (1998), found that Victorian first year drivers are
approximately three to five times more likely to be involved in a casualty road crash
than more experience drivers. Figure 9 illustrates the relationship between age and
casualty road crash risk in Victoria (when compared to the lowest crash risk group –
130
40-49 age group) (VicRoads., 2000). This pattern is also typical for drivers in other
Australian jurisdictions as well as overseas.
Figure 9: Risk of driver being involved in a casualty road crash
Maycock et al. (1991) conducted a study in Britain of drivers initially licensed
to drive at different ages (i.e. 17, 20, 25, 36 and 50 years respectively) who traveled
approximately 12 000 kilometres per year, and found that road crash risk whilst
driving solo in the first few years decreased by approximately 31% due to age and
approximately 59% due to experience. According to Gregersen and Bjurulf (1996) “it
seems that experience as well as age related factors are of vital importance. It also
seems clear that experience is of greater importance than age, at least from 17 years of
age” (p231).
5.2.2.2 The role of experience
According to an extensive review conducted by Christie (2001), both driver age
and experience (or inexperience) contributes to road crash risk and crash involvement.
Research conducted in Australia, the United States, Canada, United Kingdom and
Sweden supports the notion that as drivers accumulate greater age (along with
maturity) and experience, road crash risk decreases (Levy, 1990, Drummond and Yeo,
1992, Maycock, 1991, Mayhew, 1995, Gregersen, 1996). Compared to older drivers,
131
young drivers are at higher risk of being involved in a road crash (Williamson,
2003a). Until the age of 60, when declines in functional ability actually increase, both
increasing age and experience contribute to reduced road crash risk.
Road safety experts agree that there are two main reasons for the over-
representation of young drivers in road crashes: age and inexperience. Gusfield
(1991) proposes that road safety researchers centre of young drivers being willing to
take risks, and that it is necessary to consider what the motor vehicle represents to
young drivers and how their leisure time is spent. According to Gusfield (1991), the
motor vehicle signifies adulthood, autonomy from parents and school, as well as
being a status symbol (to opposite sex). Christie (2001) argues that “youthfulness and
inexperience tend to run in parallel for most new drivers” (p5). Experience-related
factors for young drivers, according to Mayhew and Simpson’s (1995) review of
driving experience, are far inferior in the novice driver compared to experienced
drivers. Deery (1999) proposes that a reason for the complexity of the young novice
driver road crash problem is that the task of driving is itself extremely complex.
Although young drivers learn quickly the road rules and how to handle a vehicle
(Hall, 1996), young drivers have limited experience to develop the complex, higher-
order perceptual and cognitive skills required to safely interact with the driving
environment (Deery, 1999). Critical to their road crash involvement, the skills that
young drivers often fall short on include:
• Hazard perception (i.e. detecting, recognising, and dealing with traffic
hazards);
• Attentional control (i.e. attending to the right things, in the right amounts, at
the right time);
• Timesharing (i.e. dealing with changing workloads); and
• Calibration (i.e. matching one’s performance with task demands).
(Deery, 1999, p228)
The distinction between age and experience has been referred to as ‘driving
style’ (or behaviour) and ‘driving skill’ (or performance) (Deery, 1996, Elander,
1993). Driving skill is concerned with limitations of performance on aspects of the
driving task, and is expected to improve with practice or training (although some
researchers propose that training is not beneficial to young drivers). Whereas driving
132
style is concerned with decision-making aspects of driving, such as the way drivers
choose to drive or driving habits developed over time (Deery, 1999). Driving speed
and following distance, are two such examples of driving style (Deery, 1999).
5.2.2.3 Driving at level crossings
Although younger drivers are not over-represented in level crossing collisions
data in Australia, there have been some highly publicised collisions involving younger
drivers at level crossings. Preliminary research with young drivers through use of
focus group discussions and one-on-one interviews revealed that metropolitan drivers
have very poor knowledge of protection systems other than active crossings that have
boom gates and/or flashing lights, with approximately one quarter of participants
stating they didn’t know about passive crossings (that only have a Stop or Give-way
sign). Also of concern with this younger group was the low level of knowledge of the
meaning of road markings (such as yellow painted hatching across the tracks) at level
crossings. Additionally, some younger driver hold the belief that flashing lights are
‘cautionary’ lights (like amber light at traffic lights) and that drivers don’t need to
stop when lights are activated at a crossing. Overall, younger drivers stated that they
believe they are at low risk of being involved in a collision. As such, their
performance and behaviour with regards to driving at level crossings warrants
investigation.
5.2.3 Older drivers
Increasing concern is centered on the high crash risk of older drivers. Although
older drivers currently represent only a small number of accidents, the older
driver problem is highlighted when accident statistics are presented per
distance traveled.
(Andrea et al., 2001, p1)
In comparison to younger drivers, Charlton, Oxley, Fildes and Les (2001b)
suggest that when distance is taken into account, older drivers behave similarly to
novice drivers in terms of risk. The documented road crash risk of older drivers,
particularly in the 80 plus years age group, is significantly higher than the overall
average (Di Stefano and Macdonald, 2003). However, the elevation in crash risk does
133
not automatically represent collision risk, because older drivers tend to have greater
physical frailty therefore when they do crash are likely to have injuries (Di Stefano
and Macdonald, 2003). It is indisputable that older drivers’ performance tends to
deteriorate significantly with increasing age, and there is accumulating research
(Daigneault, 2002, Stutts, 1998, Lundberg, 1998) supporting the view that a major
causal factor in this deterioration is cognitive decrements related to the ageing process
(Di Stefano and Macdonald, 2003). However, Eby and Kostyniuk (1998a) and Janke
(1994) propose that the functional impairments of older drivers do not necessarily
reduce their ability to drive safely.
5.2.3.1 Overview of the ageing driving population
Older people constitute the fastest growing sector of the driving population and
are believed to represent a high risk to road safety, given their high crash rate
per distance traveled.
(Wood, 2002, p214)
Predictions by Young (1990) indicate that the proportion of older Australians is
likely to double between 1990 and 2030 from 12% to 25%. This prediction of
Australian older people is mirrored in Japan, the United States and Western Europe
(Fildes, 1997). This increase in older people in the population extends to the driving
population. It has been suggested that “tomorrows older generation is much more
mobile than the current generation as cars and travel have become more achievable
and acceptable” (Fildes, 1997, p9). Driving for people aged 60 plus years is often
equated with mobility, and perceived as necessary to maintaining independence,
autonomy and self-esteem (Carp, 1988). According to Fildes (1997) there is evidence
to suggest that loosing the ability to drive is coupled with increased depression and
dependence on others. However, given that the elderly are continuing to drive well
into their eighth and ninth decades of life (Jette and Branch, 1992), the increase in this
driving population has important implications for road safety. Shaheen and Niemeier
(2001) suggest that there are “important consequences associated with an increased
reliance by the elderly on the automobile” (p156). As such, road crash rates create a
potentially significant problem.
134
5.2.3.2 Crash statistics and patterns of older drivers
Older drivers are generally perceived to be cautious drivers, however road crash
statistics indicate that older drivers (aged 70 + years) have a higher serious injury road
crash risk than all other age groups (including younger drivers). While it is known
that older drivers have relatively few road crashes, they are much more likely to be
severely or fatally injured given crash involvement (Fildes, 2004, Charlton, 2001a).
When distance is taken into account, the risk of a road crash for an older driver is
similar to that of novice drivers (Charlton, 2001a, Daigneault, 2002). Research
conducted by Fildes et al (2001) established projections of the future road crash risk
for the older driver population, taking into account driving behaviour, population
migration, personal wealth and health, infrastructure and technological impacts
(Fildes, 2001). This projection is illustrated in Figure 10. This research predicted
that without active road safety intervention, there would be an overall three-fold
increase in fatal road crashes involving older drivers (Fildes, 2001). By 2005, this
research indicated that there would be an increase of 261% of males and 336%
increase for females in older driver fatalities.
Figure 10: Projected older driver fatalities in Australia, 1995-2005.
Older drivers are typically involved in multiple vehicle road crashes and road
crashes in complex driving situations (Wood, 2002, Andrea et al., 2001). They are
involved in a disproportionate number of road crashes involving intersections, with
failure to stop, inattention and turning across traffic (McGwin and Brown, 1999a,
135
Ryan, 1998) and are commonly found ‘at-fault’ in road crashes (Preusser et al., 1998).
According to Drakopoulos and Lyles (2001), the “intersection environment presents
one of the greatest challenges for driver mental capacity due to the presence of
conflicting vehicular and pedestrian traffic movements, and the need for quick
decision-making in response to signs, signals, other drivers’ actions….” (p87). A
recent Australasian study of ‘blackspot’ road crash sites of older drivers found that the
principal problem for older drivers was selecting safe gaps in conflicting traffic,
manifesting itself at intersections controlled by stop or give-way signs, or traffic
signals (Fildes et al., 2000). World-wide road crash data supports this research
(McGwin and Brown, 1999b, Benekohal et al., 1994). Charlton et al (2001a) suggests
that a higher proportion of older drivers’ travel time is spent in high risk environments
(urban areas rather than on freeways or rural areas), which places them at higher
average risk per unit time or distance driven, compared to their younger counterparts.
However, the types of road crashes are different between those of younger drivers.
Older drivers are more likely to be involved in crashes at lower speeds, during
daylight hours, on dry roads, and be well under the legal BAC limit (Fildes et al.,
1994, Jette and Branch, 1992). Additionally, those aged 75 years and over
substantially exceed any other driver age group in terms of casualty road crash risk
(Diamantopoulou, 1996). International statistics show similar trends to research
conducted in Australia.
5.2.3.3 Physical ability and cognitive performance
The deterioration in driving performance in older drivers is multi-factorial,
however physical ability and cognitive performance play pivotal roles. The effects of
physical and cognitive deterioration with age has been “universally acknowledged and
widely studied” (King, 2004, p8). Physiological changes associated with ageing have
been found to affect perceptual, motor and cognitive abilities during the task of
driving (Eby, 1998b, Daigneault, 2002, Lundberg, 1998, Stutts, 1998). Shaheen and
Niemeier (2001) conducted an extensive literature review to accumulate quantitative
details and statistics concerning changes in physical factors in five key areas of
impairment: (1) vision, (2) hearing, (3) cognitive response, (4) cognitive attention and
memory, and (5) physical strength. Oxley et al (1995) added to this review and
provided more specific areas of degeneration:
• Visual acuity;
136
• Attention capacity;
• Contrast sensitivity;
• Cognitive processing ability;
• Visual field loss;
• Decision time deterioration;
• Dark adaptation and glare recovery;
• Loss of memory capacity;
• Auditory capacity;
• Neuromuscular and strength loss;
• Perceptual performance;
• Postural control and gait changes;
• Motion perception; and
• Reaction time.
(p24)
However, vision is the most recognised age-related physical decline associated
with driving performance. Vision is an important factor in driving a vehicle, with
visual impairment becoming significantly more prevalent with increasing age (Attebo
et al., 1996) through both the normal ageing process as well as the increased
prevalence of eye disease (Wood, 2002). Age-related visual impairment (such as
declines in visual field, visual processing speed, visual search, low light sensitivity,
dynamic and static visual acuity, perception of angular movement and movement in
depth, resistance to glare and contrast sensitivity, have all been found to be associated
with varying degrees of increase in road crash risk (Shinar and Scheiber, 1991, Kline
et al., 1992, Ball et al., 1993). The ageing process results in yellowing and cloudiness
of the crystalline lens, a decrease in pupil size and alterations in the integrity of the
macular pigment and neural pathways (Wood, 2002, Weale, 1992). Haegerstrom-
Portnoy, Schneck and Brabyn (1999) suggest that these changes lead to the reductions
in light sensitivity, increased glare sensitivity and reduced visual acuity. According to
Simms (1985), nearly 90% of the information translated to a driver is visual, with the
efficiency of the person’s visual perception skills likely to influence the driver’s
competence on the road.
A recent study in Queensland indicated that there is little correlation between
self-rated driving ability and actual performance (Carberry et al., 2004). This study
137
specifically examined the relationship between eye disease and driving performance.
Results suggest that drivers with visual impairment perform worse at driving tasks
such as obstacle avoidance and sign detection. This has serious implications for
complex driving environments such as level crossings. With many older drivers
being unaware of their visual limitations and the extent to which the ageing process
impairs the visual field and visual acuity, informing older drivers of their visual
limitations may assist them to modestly adapt their driving behaviour (Ball and
Owsley, 1991). However, the role that self-regulatory behaviour plays in reducing
road crash risk is a topic of great debate.
5.2.3.4 Self-regulatory driving behaviour
The processes involved in self-regulation and the factors that influence the
adoption or avoidance of self-regulatory behaviours are complicated and not
well understood.
(Charlton, 2001a, p2)
There is much speculation that older drivers self-regulate their driving
behaviour in response to a functional impairment or difficulties they experience when
driving. This speculation includes the skill in regulating driving according to their
own ability to reduce the incidence and severity of a collision. Some research
evidence advocates that caution and conservativeness prevail when it comes to
driving practices of older road users (Eberhard, 1996, Winter, 1988). According to
Charlton et al (2003), such examples of self-regulatory behaviour in the literature has
included:
• Driving more slowly;
• Traveling shorter distances;
• Making fewer trips;
• Avoiding difficult driving conditions (night-time, peak travel times and other
stress-inducing situations);
• Preferences for longer time gaps when turning or merging; and
• Avoidance of simultaneous activities while driving.
Charlton et al’s (2003) study of 656 drivers aged 55 years and older (including
29 former drivers) in Victoria, examined the extent and nature of self-regulatory
138
driving behaviour and the characteristics of those who self-regulate and those who do
not. The results indicated that self-regulators tended to be female, aged 75 plus years,
with vision problems, with arthritis, lower ratings of speed of decision-making for
safe driving, not the principal driver and not married (Charlton et al., 2003). Charlton
(2003) notes:
While it is likely that many older drivers adjust their driving adequately, it is
also possible that some fail to self-regulate appropriately and, as a
consequence, may be at higher risk of crash involvement. There is some
evidence to suggest that older drivers, or at least some older drivers, do not
self-regulate adequately. Indeed, if self-regulation was entirely successful, crash
statistics would not show an over-representation of older drivers in serious
casualty crashes.
(p9)
One notable finding from this research is that road crash risk was not
necessarily related to lower crash involvement (Charlton et al., 2003). Another study
supporting self-regulatory behaviour of older drivers found that older drivers with
small reductions in spatial vision (particularly acuity in the presence of glare and
binocular deficits) recognised their limitations and restricted their driving to the day
(West, 2003). Contrary to such findings of self-regulatory behaviour, reports on older
drivers’ self-assessment of their driving ability indicate that older drivers both under-
estimate the risk of being involved in a road crash and over-estimate their ability to
handle the driving task and feel they have total control to avoid road crashes (Brainin,
1980, Matthews, 1986). However, self-regulation (or the lack of it) not only involves
older drivers, but people of all ages. A study by Rothman, Klein and Weinstein
(1996) found that people of all ages are poor at recognising the relationship between
their actions and potential driving risks. This study also found that many people
perceive themselves as less likely than their peers to suffer harm (Rothman, 1996).
5.2.3.5 Medical conditions
Medical conditions and their association with road crash risk has also been the
source of debate. Although there has been little rigorous research examining the
impact of medical conditions on driving ability or road crash involvement, some
research has indicated that medical conditions such as diabetes, cardiac conditions
139
(Koepsell, 1994), a history of falls, kidney problems or stroke (Lyman, 2001)
increases the likelihood of an older driver being involved in a road crash. However,
as Fildes et al (1997) point out, evidence of crash associations with medical
conditions “is not always conclusive and the mechanisms not well understood” (px)
and that the “extent to which drivers with these conditions self-regulate their driving
will also influence their crash involvement” (px).
5.2.3.6 Driving at level crossings
As can be seen, there are many sensory, cognitive and physical functions that
deteriorate as people age which may ultimately impair their driving ability at place
them at high risk of being involved in a crash. Although this review did not attempt be
an exhaustive review of the effects of ageing on driving performance, it has
highlighted some important factors: (i) there is little correlation between self-rated
driving ability and actual performance; (ii) drivers with visual impairment perform
worse at driving tasks such as obstacle avoidance and sign detection; and (iii) the self-
regulatory behaviour of older drivers is a domain that is not entirely understood.
Preliminary research with drivers aged 60 years and over through the use of
focus group discussions and one-on-one interviews revealed that older drivers vary in
their perception of risk, with half of the participants stating that they are not at risk of
being involved in a level crossing collision, while the other half stated they are at
some degree of risk. Participants that stated they are not at any risk whilst driving at a
level crossing referred to using self-regulatory behaviour such as avoiding certain
‘risky’ crossings or becoming familiar with crossings they use regularly. Those
participants that stated they may be at risk whilst driving at a level crossing cited that
inattention or poorly designed crossings could be significant factors. Pressure from
other drivers was also a common theme that emerged from this preliminary research.
Although most older drivers indicated that they don’t take risks at level crossings,
many declared that they have had a close-call at a level crossing and many stated
visibility at night-time was a particular concern for them. Of particular concern with
this older driver group, was that the majority of urban participants indicated that
passive crossings (i.e. Stop or Give-way signed crossings) were ‘only in the bush’.
With the results of this preliminary research as well as the fact that older drivers are
over-represented in fatal collisions at level crossings, understanding why ageing
drivers are at greater risk of such collisions is of key important for road safety.
140
5.2.4 Heavy vehicle drivers
5.2.4.1 Overview of the safety in the heavy vehicle industry
Long distance trucking makes an often under-estimated but very substantial
contribution to the Australia economy and society…. These achievements of the
long distance trucking industry come with a cost. These costs include safety
problems facing drivers and the general motoring public.
(Quinlan, 2001, p33)
Although the past decade has seen a decline in fatal road crashes involving
articulated trucks (long haul), incidents still account for a significant and
disproportionate number of all road fatalities (Quinlan, 2001). It must be noted
though, that in only about 25% of such fatal road crashes, the truck driver is held
responsible. Nevertheless, driving a truck for a living remains one of the most
dangerous occupations.
There are a variety of ways in evaluating safety in the long haul trucking
industry, including indications of the hazardous nature of this industry. From a
financial point of view, there are two distinct factors:
• Claims on Compulsory Third Party (CTP insurance) have escalated for this
industry during the past decade; and
• Comprehensive insurance have seen commercial vehicle insurance rates rise
dramatically (in some cases by more than 30%).
(Quinlan, 2001)
The alternative view - that of comparing road crash statistics to other transport
industries - indicates that despite funds being expended on upgrading roads than
maintaining an ‘ageing rail network’, the safety performance of the trucking industry
is be far worse than the main alternative of land transport (that of rail) (Quinlan,
2001).
5.2.4.2 Crash statistics and patterns of heavy vehicle crashes
In 1995, the Federal Office of Road Safety (now Australian Transport Safety
Bureau) commissioned analyses on the nature of urban truck road crashes in Australia
(Sweatman et al., 1995). A large body of work was undertaken for these analyses
including a literature review, mass data analysis, detailed post-crash analysis of fatal
141
crashes, analysis of accident black spots and in-depth investigation. The findings
from these analyses include:
• There were around 1000 serious heavy vehicle crashes per year in urban areas
costing the community approximately $100 million per year;
• Serious crashes formed 50-75% of serious rigid truck crashes and 25-50% of
articulated truck crashes;
• Majority of causalities were drivers of cars involved in crashes with heavy
vehicles;
• Majority of fatal truck crashes occur in non-urban areas;
• Non-urban truck crashes are significantly worse than urban crashes;
• Most fatal truck crashes occur in lower speed zones; and
• Precipitating factors in fatal crashes included inappropriate behaviour,
inattention, disregards of traffic controls and excess speed (often on the part of
the other vehicle or pedestrian).
Table 8 provides a summary of the annual number of fatal road crashes as
recorded by FORS. During this period (1981-1998), it is evident that there was a
downward trend in the total number of road crashes and fatalities involving all
vehicles, however this trend is not observed in fatal road crashes involving vehicle
vehicles. Since 1989, there is an observed significant improvement in both fatal road
crashes involving trucks and other vehicles. Overall, the safety performance of
articulated trucks has shown a considerable improvement since the late 1980’s.
However, trucks remain over-represented in fatal road crashes (whereby the rate of
involvement in fatal road crashes is compared to the proportions of vehicles
constituted by heavy trucks) and account for a significant component of the national
road toll (Quinlan, 2001). As noted earlier, most persons killed in fatal road crashes
involving heavy vehicles are not truck drivers, but rather members of the public
(Quinlan, 2001). According to an inquiry of safety in the long haul trucking industry
(Quinlan, 2001), this safety issue adds a fundamental public safety dimension in
reviewing the safety performance of this industry outside that of truck driver safety.
142
Table 8: Fatal road crashes and fatalities, Australia 1981 to 1998
Source: Quinlan, M. (2001).
In comparison to other countries, the Australian fatal truck road crash rate per
kilometre traveled is 47% greater than that of the U.S.A., 39% higher than the U.K.,
comparable to both Germany and Canada, 20% lower than Sweden, 45% lower than
France and 55% lower than New Zealand (Haworth et al., 2002). However, these
researchers (Haworth et al., 2002) propose that the higher fatality rates observed on
Australian roads compared to the U.S.A. and U.K. may largely be explained by the
lower proportion of truck travel on divided and limited access roads in Australia, as
well as possibility speed limits.
5.2.4.3 Speeding
There are grounds to believe that the link between speed and safety outcomes
may be even more critical for heavy vehicles than for light vehicles.
(National Road Transport Commission, 2004, p16)
143
According to a report released in 2004 by the National Road Transport
Commission (2004), concern and action on heavy vehicles has largely focused on the
small percentage of drivers that exceed the speed limits by a substantial margin.
However, this report claims that although those drivers represent a significant safety
hazard, low level speeding (by just a few km/hour) also involves substantial risk
(National Road Transport Commission, 2004). ‘Low level’ speeding is asserted to be
important to overall safety outcomes because it is far more prevalent than extreme
speeds (National Road Transport Commission, 2004). To measure compliance of
heavy vehicles over 12 tonne with speed limits, all such vehicles are required to be
fitted with a speed limiter (designed to limit peak highway speeds to 100km/hr)
(National Road Transport Commission, 2004). However, speed distribution data
indicates that there are many non-compliant truck drivers on the road. ARRB
Transport Research conducted a study for Austroads on heavy vehicle compliance
with speed (George, 2002b) with six years of data (years 1995-2000) and 45 million
observations. This investigation found that for aggregated Australian data (excluding
Western Australia and Northern Territory) averaged over the years 1995-2000, 17%
of the observed class 3 vehicles and 26% of the class 9 vehicles were detected
exceeding the posted speed limit (i.e. 100km/hr) (George, 2002b). Figure 11 shows
the cumulative frequency plot for class 3 vehicles (two axle rigid vehicles) observed
exceeding the posted speed limit (George, 2002b).
Figure 11: Austroads class 3 vehicles degree of speeding
144
5.2.4.4 Fatigue and sleep deprivation
Fatigue has been shown to affect mental alertness, thereby decreasing an
individual’s ability to operate a vehicle safely and increasing the risk of human error
that could lead to fatalities and injuries. Drowsiness slows reaction time, decreases
awareness, and impairs judgment. Driver impairment due to drowsiness is known to
be a major contributing factor in many crashes involving commercial-vehicle drivers.
(Hanowski et al., 2007, p1)
Fatigue and sleep deprivation have long been recognised as being critical safety
issues in the trucking industry, with driver fatigue being found to be a major cause of
long-haul truck road crashes (Haworth et al., 1988, National Transportation Safety
Board, 1995). The two most important factors that are likely to increase fatigue
include: long periods of exposure to a monotonous task such as driving (Hamelin,
1987, Krueger, 1989) and driving in the early hours of the morning (Feyer et al.,
1995). According to the ‘Fatigue Expert Group’ (2001) there are considerable
incentives in the economic and social profile of the transport industry for scheduling,
trip planning and consequent driver practices that increase the risk of fatigue for the
driver (Economic Associates Pty Ltd, 2002). Additionally, competitive pressures,
payment systems, contracting arrangements and even the unintended consequences of
the driving hours regime combine to create an environment in which fatigue has
become an accepted part of industry practice (Economic Associates Pty Ltd, 2002,
p5).
However, in recent years, there have been a large number of changes in the
long-haul trucking industry in Australia, with many of the changes having an impact
on fatigue management practices. In 1998 Queensland’s Department of Transport
(Queensland Transport) implemented the National Driving Hours Policy to better
manage heavy vehicle driver fatigue. The policy was introduced through the
Transport Operations (Road Use Management - Fatigue Management) Regulation
1998 and makes use of log books and prescriptive driving, and work and rest limits
(Queensland Transport, 2007). An evaluation of this program was conducted over a
six year period (1996-2002), with findings indicating that drivers are now less likely
to report speeding to meet a deadline, feeling tired, and experiencing difficulty
concentrating (Burgess-Limerick and Bowen-Rotsaert, 2002). Other jurisdictions in
Australia have implemented the National Driving Hours Policy (New South Wales,
145
Victoria, South Australia and Tasmania) which prescribes maximum working hours
for heavy vehicle drivers on a daily and weekly basis, including provision for short
and long work breaks (Economic Associates Pty Ltd, 2002). This regulation also
prescribes requirements for the maintenance of logbooks for recording of driving
working hours (Economic Associates Pty Ltd, 2002). Western Australia and the
Northern Territory are the only two jurisdictions in which this regulation does not
apply. These two jurisdictions have introduced Fatigue Management Codes of
Practice under Occupational Health and Safety legislation. In the Australian Capital
Territory, driving hours are not regulated. However, such regulations have been
criticised for focusing on a symptom of poor fatigue management (i.e. hours actually
worked). In their report for the National Road Transport Commission, the Australian
Transport Safety Bureau and the New Zealand Land Transport Authority, the ‘Fatigue
Expert Group’ (2001) concluded however that the prescribed driving hours do not
account for circadian patterns/time of day factors (especially night work). It is
therefore evident that fatigue management practices in multifaceted economic and
social environments such as the long-haul trucking industry are fraught with continual
barriers in their implementation and will no doubt receive repeated criticism.
5.2.4.5 Drug use
Drug use within the heavy vehicle industry is not within the scope of this
research program, however it is important to acknowledge that there is a clear
association between stimulant use by truck drivers and road crash risk (Swann, 2002).
Drug use by truck drivers is a great public concern for two reasons. Firstly, it has
been said that heavy vehicle traffic is expected to increase 75% in the period to 2010
and secondly truck drivers that test positive to stimulants have an crash risk similar to
car drivers who test positive to a BAC of 0.10 – 0.15 (such levels of alcohol represent
gross impairment) (Swann, 2002). Therefore, acknowledging the problem that drug
use poses to truck safety and other road users is important.
According to the Victorian Parliamentary Road Safety Committee (1996), it is
well documented that use of various types of illegal drugs is highly prevalent in the
general population. At the core of drug use is the intent to alter physical and/or
mental function, which is also capable of impairing driving skills (Marowitz, 1995).
Research conducted by Drummer (2002) supports this assumption. Controlled studies
reveal that drug use contributes to the physiological and mental impairment of the
146
driving task. Of most concern to road safety experts are the debilitating effects of
drug use on driving skills and behaviour and the prevalence of drug use in the driving
population (Ramsay and Prem, 2000). The drugs that are of particular interest to road
safety researchers include cannabis and the central nervous system stimulants. These
drugs are extensively reported as the ‘drugs of choice’ by heavy vehicle drivers that
have been found to drug drive (Victoria Police State Highway Task Force, 1995,
Hartley and Arnold, 1996, Swann, 2002). Quinlan (2001) states:
The use of drug stimulants by truck drivers to combat fatigue has long been a
feature of the long distance trucking industry in Australia (since at least the
1970s).
(p77)
However, Quinlan (2001) suggests that “given the illegality of many stimulant
drugs, obtaining accurate information on the extent and nature of the practices has
always been difficult. It is therefore impossible to discuss the use of drugs with
complete precision” (p78). However, surveys of truck drivers suggest that there is a
consistent pattern whereby drug use is widespread within the long-haul trucking
industry (Quinlan, 2001). Enforcement practices largely rely on on-road detection
which represents a time-consuming task by police and is only like to detect some of
the most extreme cases (Quinlan, 2001). Quinlan (2001) also proposes that there has
also been evidence of the supply of stimulants by employers and the removal of
incriminating evidence by tow truck operators in the event of a collision.
5.2.4.6 Sharing the road
An inquiry coordinated by the Motor Accidents Authority of New South Wales
to investigate safety in the long-haul road transport industry, found that specific
mention was made of the fact that some car drivers expect trucks to have similar
performance characteristics to cars and motorists in general appear to have received
no indication of the significant differences as part of their driver education (Quinlan,
2001). This inquiry found a strong case for including information about heavy vehicle
characteristics (e.g. signage, braking, turning) in the formal training of general
motorists as part of the licensing process (Quinlan, 2001). The inquiry formed the
view that:
147
… inadequate understanding of heavy vehicles amongst other road users is an
issue that needs to be addressed both because it exacerbates safety risks to
other road users and places additional unnecessary pressure of truck drivers.
(Quinlan, 2001, p83)
This issue of ‘sharing the road’ has also received great attention in the U.S.A.
Allocation of funding to educate the motoring public on how to share the road safely
with commercial motor vehicles is part of the ‘Share the Road/No-Zone Campaign’
which works in conjunction with regulatory agencies to increase the public’s
recognition of truck and bus limitations. Although this program can assist in
educating motorists about certain areas around a truck where the truck driver cannot
see motorists, it is not designed to influence the perception or beliefs about trucks by
motorists.
5.2.4.7 Driving at level crossings
Investigations by the ATSB of fatal collisions at level crossings involving heavy
vehicles indicate that for the majority of collisions, the collision was a consequence of
motorist behaviour (either through error, lapses or violations). Although more than
one contributing factor have generally been found in each of the investigations at
level crossings, ‘failing to stop’ by the truck driver has been cited in numerous
investigation reports.
Behaviour that is common among many drivers is a rolling stop. By not coming
to a complete stop, drivers have less time to notice the approach of a train. In
addition, checking for a train while the vehicle is still moving can make the
remaining distance too short to be able to stop.
(Coghlan, 1997)
Additionally, familiarity, complacency, expectations of trains and visibility
problems (such as sighting distances) have also been cited as contributing to truck-
train collisions.
148
5.3 QUALITATIVE RESEARCH WITH TARGET GROUPS
5.3.1 Objectives
Specifically, the research explored the following issues:
• Knowledge of road rules;
• Exposure to different protection systems;
• Awareness of collisions and near-misses;
• Design of crossings;
• Familiarity of crossings;
• Importance of the problem in relation to other road safety issues;
• Attitudes, beliefs and perception of risk;
• Effectiveness of enforcement and penalties;
• Cognitive ability to judge train speed and stopping distances; and
• Constructs of message acceptance and delivery.
5.3.2 Method
5.3.2.1 Ethical clearance
Ethical clearance for data collection for this formative research phase was
gained from the Queensland University of Technology Human Research Ethics
Committee (QUT Ref. No. 3550H).
5.3.2.2 Sampling
A non-random sampling technique was used in this formative research. The two
types of non-random sampling techniques used were purposive and convenience
sampling. The purposive sampling technique was used primarily for focus groups
discussions. A purposive sample is one that is chosen by the researcher subjectively
and attempts to obtain a sample that is representative of that population (Tilley, 1990).
Although this type of sampling allows information to be obtained from the target
population, it is likely to overweigh sub-groups in the population that are more readily
accessible. Convenience sampling technique was used for semi-structured interviews.
A convenience sample also has limitations. Although convenience samples comprise
of participants who are available in a convenient manner to researchers, there is no
randomness (Tilley, 1990). Thus the likelihood of bias is considered somewhat high.
149
However, these two types of sampling techniques were the most appropriate and
practical within the budget and resources of the project.
This combined approach (purposive and convenience sampling) was seen as an
appropriate and practical data collection strategy within the qualitative and
exploration methodology. Focus groups offer the opportunity to explore and gather
rich data on a specific issue and gain insights into the experiences and perceptions of
its participants. One major advantage of this data collection tool is that data can be
collected from numerous people in a relatively short period of time (Beyea, 2000).
Semi-structured interviews is one of the most frequently used qualitative methods,
and allows for focused, conversational, two-way communication to obtain a range of
insights on level crossing safety. These interviews used a script, consisting of a set of
questions as a starting point to guide interaction with the participant (Sampson, 1972).
Nevertheless, as the aim is to capture as much information as possible on the
participant’s views about level crossing safety, the interview follows the participants
thinking and poses new questions from answers given by the participant (Sampson,
1972). Consequently, each interview can be quite different from other interviews
(Drever, 1995, Wengraf, 2001).
Both focus groups and semi-structured interviews were conducted with
participants from each of the three road user groups in both urban and regional areas
in Queensland by trained researchers in road safety. Locations to conduct the focus
groups and interviews were selected to utilise settings that were familiar, practical and
non-threatening. Older drivers were mainly recruited through existing social and
charity organisations. In the metropolitan area, bowls clubs and social organisations
were approached and given flyers and posters advertising an invitation to take part in
a discussion on road safety at the organisation’s premises. In the regional areas, the
charity ‘Meals on Wheels’ and the ‘60 and Better’ program were approached and
were also sent posters and flyers for the sessions. The criteria for participation for the
older group included:
• Aged 60 years or over; and
• Have a valid driver’s licence.
Younger drivers were recruited through the education facilities of Universities
and TAFE colleges in both regional and urban areas within Queensland. Approval
was sought from both of these educational facilities to conduct either focus groups or
150
interviews with students on their campuses. Focus groups in the urban area were
advertised on first year student’s noticeboards. Through arrangements with the
Course Coordinator of one of the University’s departments, students were given
academic credit for attending the sessions. Interviews in the urban area were
conducted through convenience sampling on the campuses of a large TAFE college
and as well as a University where focus groups were also conducted. Students were
approached in public areas by the interviewers and invited to participate in an
interview. These participants were given a gift of $10 for their time. In the regional
areas, interviews were conducted through convenience sampling on the campuses of a
TAFE college and a University in central Queensland. Students were approached in
public areas by the interviewer and invited to participate in an interview. These
participants were also given a gift of $10 for their time. The criteria for participation
for the younger group included:
• Aged between 17 to 24 years; and
• Have a valid learners’ permit or driver’s licence.
Truck drivers were recruited through convenience sampling at a large
‘Truckstop’ in Brisbane. Assistance was sought from the Queensland Trucking
Association (QTA) to gain permission from the Truckstop to conduct interviews on
site. On approval from the Truckstop, two trained interviewers approached drivers on
their meal breaks and invited them to participate in an interview. The interviews were
conducted over a period of two days. All participants received a travel mug as a gift
for their time. The criteria for participation for the heavy vehicle group included:
• Have a valid heavy vehicle licence; and
• Drive a heavy vehicle.
5.3.2.3 Procedure
Information sheets and consent forms were provided to each participant prior to
commencement of focus groups and semi-structured interviews. Participants were
informed that participation was purely voluntary and that their responses would
remain anonymous through de-identification of any collected data. Consent forms
were signed by participants prior to commencement of the discussion/interview.
During data gathering, verification and validation of responses was strengthened
through the use of interviewer probes, follow-up questions and reiteration. For the
151
focus group discussion, the researcher’s aim was to facilitate discussion between
participants to ensure dominant participants didn’t control the discussion. During
each focus group, an inductive approach was used, which allowed the researcher to be
flexible in exploring issues and themes as they arose. To allow external validation,
focus group discussions were tape recorded with permission from participants and
subsequently transcribed. Transcribed data from these tapes will be de-identified to
protect each participant’s identify. The focus group sessions took approximately 60
minutes to administer, whilst interviews took approximately 20 minutes.
5.3.2.4 Materials
An agenda for the focus group session was developed in accordance with
guidelines for conducting focus groups. This agenda included: welcome, review of
agenda, review of goal of the meeting, review of ground rules, introductions,
questions and answers, and closing comments. The moderator’s guide was
synonymous with the interview schedules (although within the group setting greater
flexibility was required to allow the group process to occur) and was developed
specifically for each target group. Questions were developed with input from experts
in the area, and were based on the themes identified from Study One (train drivers and
expert panel participation). The focus group moderator’s guide was developed with
the interview schedules, and were piloted and re-drafted, with the final guides
designed to gather data on:
• Demographic information;
• Exposure to level crossings;
• Knowledge of level crossing types, road rules and safety information;
• Attitudes and beliefs to driving at level crossings;
• Self-reported behaviours;
• Behaviours of other motorists; and
• Road safety message acceptance and delivery.
5.3.2.5 Data analysis and reportage
Raw data collected from the focus group discussions was collected in a
relatively unstructured form (by use of tape recordings) and then transcribed. The data
was then analysed through a qualitative thematic analysis process (Wengraf, 2001).
Although the moderator’s guide was designed to gather information on a range of
152
topics, reportage is broken down into key themes that emerged from the data. These
themes provide the framework for the categorisation of the responses, with
participant’s own words being identified in the text by the use of italics. There were
five major themes that consistently emerged throughout the data from the younger and
older driver groups, while six themes emerged from the truck driver group. Results of
this thematic analysis are presented according to these themes, with regional and
urban responses being integrated (older and younger drivers only). Following
presentation of these key themes, a summary of the key differences between regional
and urban responses is summarised.
5.3.3 Results
5.3.3.1 Sample characteristics
A total of 122 participants were involved in this formative research phase, with
70 participants (57.4%) taking part in the semi-structured interviews and 52 (42.6%)
in the focus group sessions. In total, there were 51 females (41.8%) and 71 male
(58.2%) involved in this formative research phase. Table 9 illustrates data collection
from each of the road user groups in both regional and urban areas.
Table 9: Data collected during formative research phase Target Group Component of Data
Collection Location n
Heavy vehicle drivers (all areas)
Semi-structured interviews
Major truck stop 26
Young drivers (urban)
Semi-structured interviews and focus groups
Queensland Transport Licensing Centre and University
25
Young drivers (regional)
Semi-structured interviews
TAFE and University 28
Older drivers (urban)
Semi-structured interviews and focus groups
Senior Citizens Bowls Clubs
23
Older drivers (regional)
Focus groups Meals on Wheels 60 and Better program
20
5.3.3.2 Older drivers
There were a total of twenty-three participants in the urban group of older
drivers. Nineteen people took part in two focus groups and four participants were
interviewed, in the metropolitan Brisbane area. Ten participants were male (43.5%)
153
and 13 were female (56.5%). The mean age of the group was 76.3 years (range 61-89
years) and the mean years licensed as a driver was 50.22 years (range 17-75 years).
The majority of participants were married (n=11, 47.8%) or widowed (n=6, 26.1%)
and were born in Australia (n=22, 95.7%). About half the participants (n=11, 49.9%)
stated they traveled 50km or more from home ‘once a month’ or more.
Twenty participants took part in three focus groups in a regional town in
Queensland. Eleven were males (55%) and 9 were females (45%). The mean age of
the group was 67.99 years (range 61-85 years) and the mean years licensed as a driver
was 46.95 years (range 30-69 years). The majority of participants were married (n=9,
45%) or widowed (n=6, 30%) and were born in Australia (n=19, 95%). The majority
of participants drove a vehicle four or more days per week (n=18, 90%) and 75%
(n=15) encountered both active and passive crossings regularly (everyday or a few
times per week). Traveling 50km or more from home was quite common with 70%
(n=14) of participants reporting that this occurred ‘once a week’ to ‘once a month’.
Perception of risk
There was a mixed perception of risk among regional older participants, with
half perceiving that they are not at risk of being involved in a level crossing collision,
whilst the other half believed that there is some degree of risk. Those participants that
believed they are not at risk whilst driving at level crossings believe they follow the
road rules, they use protective behaviours such as avoiding certain crossings, or
becoming more familiar with crossing. Those regional participants that believed they
are at some risk believe inattention and poor designed crossings are likely factors that
would increase their level of risk. The majority of regional participants reported rarely
taking risks at level crossings, although many stated that other motorists are the ones
that take risks frequently at crossings.
There appeared to be a high level of awareness that collisions between trains
and motor vehicles occurs, with one regional participant stating some accurate
statistics. Near-misses were frequent with many participants in this regional group.
Many participants stated that they had close-calls with particular crossings and
therefore tried to avoid these crossings. It is a wake-up call for sure. Most of the
participants that recalled near-misses indicated that they were quite shaken by the
experience and try to be more cautious at crossings. Various reasons were suggested
as to why they were involved in a near-miss with some suggesting that complacency
154
played a big role. Some participants suggested that you look but you don’t always see
as being the reason why they experienced a close call with a train.
The majority of urban participants felt that level crossings were not high on the
safety agenda (in comparison to drink driving, speeding etc). A couple of urban
participants asked if anyone was ever killed while driving at level crossings and were
surprised to hear how many motorists are fatally injured each year in Australia. The
response given by one participant is it that many? appeared to sum up the low level of
perceived risk that this group held. However, the majority of urban participants felt
that level crossings were often complex driving environments that could place them at
risk, but they maintained that protective behaviours (such as only driving during the
day or at familiar crossings) would reduce their risk of being involved in a collision.
The belief that I drive on roads that I know well appeared to be widespread in this
urban group. Most of the urban participants stated being familiar with a number of
local crossings and thought that due to this familiarity, it made driving at such
crossings easier and safer. Unfamiliar crossings (such as those in areas where they
rarely or never driver) were believed to cause more risk to participants and many
stated that they try to avoid driving over such crossings. Some participants indicated
that they could get confused in unfamiliar environments, particularly crossings that
are complex. The notion that there are too many lights and signs going on at some
level crossings was felt by many participants in the urban group. Overall, urban
participants felt that due to low driving exposure at crossings and ensuring they obey
the road rules at crossings, that there would be an extremely low risk of being
involved in a collision with a train. Near-misses were not as common with urban
participants, with only one participant stating that they got stuck on a crossing due to
congestion ahead of the crossing. This participant suggested that it was a scary
experience.
Self-reported and intended driving behaviour
Both the regional and urban participants indicated that they obey the rules at
level crossings and that it is other drivers that put them at risk. Many participants
noted that it is the young drivers particularly that cause them greatest concern when
driving at level crossings. No participant admitted to intentionally violating or would
be willing to violate the road rules at level crossings such as driving through activated
warning systems or failing to stop/give-way to trains at passive crossings. One urban
155
participant stated that you’d have to be stupid to take risks at railway crossings, while
one regional participant stated that the hoons will always try to beat the train.
Risk factors for being involved in a collision
Although both the urban and regional older driver groups indicated that they
don’t take risks at level crossings, many participants in the urban group mentioned the
difficulties that they sometimes face when driving in such complex traffic
environments (such as at busy urban level crossings). Behaviour of other motorists
was considered to be an issue with driving safely at level crossings for the older urban
group. Most participants in this group felt that other motorists put pressure on them
by beeping the horn, yelling at them, or driving around their vehicle when they are
waiting for flashing lights to cease flashing. The majority of participants felt that
such behaviours were unnecessary and that younger drivers were largely responsible
for such behaviours. Other perceived risk factors for the urban group were similar to
those of the regional group. Confusion and distraction were discussed, with
comments that unfamiliar crossings can be confusing, roads with many turning lanes
were confusing and signs, billboards and other objects on the road were distracting.
Many participants in the urban group mentioned protective or compensatory
behaviours such as being diligent, driving in off-peak times and getting a relative to
drive in an unfamiliar environment. The majority of participants in the urban group
mentioned the difficulty they face with the ageing process and how it influences their
driving ability. Decreased vision, poorer hearing, slower reaction times and reflexes
were the predominant factors that were declared to pose problems when driving at
level crossings. When you can’t see or hear as well as you used to it makes it difficult.
Such comments were widespread among urban participants.
Among the older regional group, there was a very high level of familiarity with
local crossings and knowledge of different types of crossings. Many regional
participants indicated that they avoid driving at unfamiliar crossings as this could be
more stressful than driving at crossings that they regularly use. Poor visibility at
nighttime (due to degeneration of their vision) was also stated by many participants as
being a problem for driving at level crossings. This was a reason that many
participants stated that they avoid driving through crossings at night or when there are
low levels of light (such as dusk or dawn). Some regional participants mentioned that
their reflexes and flexibility were not the same as they used to be. Reflexes and
156
flexibility were felt to be an issue at crossings that have a sharp angle of approach
where participants have to turn to look back up the track. Many participants in this
group indicated they as they got older, driving generally became harder physically for
them. Quite a few participants also suggested that due to degenerating vision and
hearing that driving at level crossings was stressful for them and that they would
avoid situations that they were not confident driving. Like their urban counterparts,
pressure from other drivers was a common issue that emerged in this older group.
Many regional participants stated that other drivers pressure them by beeping the
horn, yelling at them, and overtaking them at the crossing. All participants felt that
such behaviours were unnecessary and the majority held the view that younger drivers
were responsible for such behaviours.
Knowledge and perceptions of laws and enforcement
There was a very high level of knowledge about the different types of protection
systems at level crossings (in both regional and urban areas), although very few
regional participants knew of the fines and penalties for breaches at level crossings.
One regional participant stated that whatever they are, they should be higher. The
majority of regional participants indicated that they rarely (if ever) see police
enforcing the road rules at level crossings and that there subsequently would be a low
risk of any motorist being caught by police. One of the participants remarked that
police have better things to keep them occupied and implied that it would be a waste
of police resources to do patrols at crossings in regional areas. Although yellow
hatching road markings (denoting ‘keep clear’) are more likely to be in urban areas,
no regional participant could correctly explain what such road markings mean, with
many participants stating that they had never seen any crossing in regional areas that
had such road markings.
The urban group did not have the same level of knowledge about the different
types of level crossings as their regional counterparts, as passive crossings (with Stop
or Give-way signs only) were thought to be only in the bush. All participants
indicated that they were familiar with active crossings, especially with their local
level crossings. Like their regional counterparts, there was generally a poor level of
knowledge about penalties and fines for breaches at crossings, with many participants
stating that they never see police enforcing the road rules at crossings. Although
many active crossings in urban areas have yellow hatching road makes that denote
157
‘keep clear’, only a couple of urban participants could state the meaning of such road
markings. Many participants claimed that they have never seen any yellow road
markings even though they regularly drove over the local crossing that had these
markings. Generally, the urban group had a high level of knowledge of the road rules
for driving at level crossings although some participants did not know that it was
illegal to queue over a crossing.
Usefulness of education and publicity
Most regional participants felt that some education about level crossing safety
was a good idea, however others could not really see the point of it as there is a low
risk of being involved in a collision compared to other areas of driving (i.e. speeding,
drink driving etc). There was a general agreement that road safety advertisements
served to reinforce rather than change their behaviour as all participants felt they were
already safe drivers. ‘Informational’ television campaigns were considered to be
effective for many participants in this group, especially when combined with visual
effects. A number of regional participants stated that it was the visual images that they
recalled when watching road safety advertisements. Many of the regional group
suggested the use of positive rather than negative messages, such as supplementing
messages with tips on how to drive safely (e.g. towing a caravan across a crossing).
Credible sources of information delivery included police, ambulance officers or train
drivers. Additional to television campaigns, the majority of participants indicated that
talk-back radio would be an effective delivery method for this group as they consider
that experts talking about the issue to be highly credible sources of information.
Public talks from police, ambulance officers or train drivers were also seen as being
valuable to this age group.
Many urban participants felt that education or publicity for level crossing safety
was not necessary as it would be a waste of time for older drivers as they already did
the right thing at level crossings. Most of the urban group believed that if there was
to be an education campaign for older drivers, then a reminder campaign about
driving safely at level crossings and changing risk perception of older drivers would
be the most effective. Like their regional counterparts, urban participants felt that
‘informational’ television campaigns would be effective for many participants in this
group, especially when combined with visual effects. Additional to television
campaigns, the majority of urban participants indicated that talk-back radio would be
158
an effective delivery method as they frequently listen to the radio and value the
sources of information that talk-back radio provides. This group of urban participants
also suggested public talks from police, ambulance officers or train drivers as being
valuable in the delivery of messages.
Differences between regional and urban group responses
The most notable differences between regional and urban older group responses
were related to risk perception and knowledge about the different types of protection
systems. Regional drivers appear to believe that they are at greater risk of being
involved in a level crossing collision with many recalling near-misses. The main
reasons given for such near-misses largely included complacency due to familiarity of
local level crossings. On the other hand, urban drivers generally held the perception
that level crossings were not as dangerous as other aspects of driving, with many
participants doubtful that motorists die at such intersections. Many participants in the
urban group also were surprised that transport authorities were investing financial
resources into level crossing education programs for motorists as the majority
believed that it is not a major road safety priority. Urban drivers also had a poor level
of knowledge about the different types of protection systems, other than active
systems with boom gates (such as in metropolitan areas), with some participants
unaware that some level crossings have no active protection system to warn motorists
of an approaching train.
5.3.3.3 Younger drivers
There were a total of twenty-five urban participants, with approximately half
taking part in focus groups (n=13, 52%) and half in interviews (n=12, 48%) in the
Brisbane area. Fourteen participants were male (56%) and 11 were female (44%). The
mean age of the group was 20.4 years (range 17-28 years) and the mean years
licensed as a driver was 2.74 years (range 0.17-10 years). The majority of participants
(n=23, 92%) held an open driver’s licence, were single (n=19, 76%) and were born in
Australia (n=23, 92%).
Of the regional drivers that took part in this research phase (n=28), ten
participants were male (35.7%) and 18 were female (64.3%). All of these participants
took part in the semi-structured interview in regional Queensland. The mean age of
the group was 19.78 years (range 17-24 years) and the mean years licensed as a driver
159
was 2.56 years (range 0.08-8 years). Three quarters of the participants held an open
driver’s licence (n=21, 75%). Nineteen participants stated they were single (67.9%)
and most were born in Australia (n=26, 92.9%).
Perception of risk
Overall, there was a low perceived level of risk among regional participants as
they felt that they are familiar with level crossings in their area. Additionally, the
majority of this group believed that at crossings with low train traffic volumes they
are less likely to be involved in a collision than at crossings with more train traffic
volumes. It there are more trains then there is more risk. The majority of regional
participants felt that they are a safe driver and would be able to avoid a collision at a
level crossing. There was a small number of participants that were concerned that
inattention, poor visibility of trains, stalling their car on a crossing, or being a
passenger (i.e. not being in control of the vehicle), could pose some degree of risk of
being involved in a collision with a train. Being familiar with a level crossing was
deemed by half of the participants as making them drive more carefully, whilst the
other half indicated that it would make them more complacent when driving at these
crossings. Male participants were more likely to perceive that being familiar with
certain crossings would make them less likely to be involved in a collision with a
train, while female participants were more likely to hold the perception that driving at
crossings on a regular basis would make them more likely to be involved in a
collision with a train due to being on ‘auto-pilot’, with one regional participant stating
that sometimes you get to your destination and can’t remember how you got there.
Urban participants also appeared to believe that there was little risk whilst
driving at level crossings. The majority of urban participants felt that they were not at
risk of being involved in a level crossing collision. Those participants that perceived
there was some degree of risk of being involved in a collision, thought it would be
due to either equipment malfunction (such as faulty boom gates/flashing lights) or
stalling their car at the crossing. Many participants in this group raised the concern
that if they stalled their vehicle on the crossing then this would be extremely
dangerous. That’s how people get killed. It became apparent from the responses from
many participants in the urban group that there is an urban myth that the majority of
collisions between trains and motor vehicles are a result of motorists stalling their car
on a crossing. This issue was raised on numerous occasions during the discussions,
160
with many participants fearing that it could happen to anyone. Urban participants that
spoke of performing high-risk behaviours (such driving through flashing lights whilst
the boom gate is descending) indicated that they felt that their chances of being hit by
a train were small. One participant commented that you have more chance of being hit
by a bus. Like their regional counterparts, younger urban participants were divided in
their beliefs about familiarity and complacency. The group appeared to be equally
divided as to whether they drive more or less carefully at crossings that they drive
through frequently. No particular gender in the urban group stood out as holding one
belief over the other.
Self-reported and intended driving behaviour
Most regional participants stated that many motorists (including themselves)
disobey road rules at level crossings and that their ability to judge the distance of a
train as justification for why they took risks. There is usually plenty of time. Many
drivers admitted to risk taking behaviours such as not stopping at Stop signs, going
through flashing lights, racing the train to avoid waiting at boom gates, following the
car in front across the passive crossing without scanning for a train, and driving
through a crossing even when poor visibility. Those regional participants that were
concerned about there being some degree of risk at level crossings were less likely to
admit to disobeying the road rules at level crossings. Waiting for a train to go through
the crossing was seen by many regional participants as really annoying, especially
with a freight train. Many regional participants stated that long freight trains (with
more than 30 wagons) would take ages to go through the crossing and therefore they
would try to beat the train to the crossing to avoid delays. This included disobeying
road rules such as not stopping at a crossing or going through the activated flashing
lights. Participants that took risks to beat a freight train didn’t view this as risk-taking
as they believed that they had sufficient time to cross in front of the freight train.
Many urban participants also admitted to risk taking behaviours such as driving
around boom gates, queuing over crossings, racing the train to avoid waiting at boom
gates, going through flashing lights, and not stopping at Stop signs. The majority of
these participants that admitted to such risk taking behaviours indicated that this risk
taking was less dangerous than risk taking such as drink driving, speeding or driving
through red lights at an intersection. More people die drink driving than being hit by
trains…surely. Justifying their risk taking behaviours was common in the urban
161
group, with many participants stating that ‘no one gets hurt’ when they disobey road
rules. Some participants stated that they frequently disobey the rules at level crossings
and have never had a near-miss with a train. The reasons for such justifications
included: infrequent trains, being able to judge the distance that the train is from the
level crossing, and that trains will slow down if they observe a car at a crossing.
Although most urban participants didn’t encounter passive crossings in rural or
remote areas, there were many participants that stated they would be willing to violate
road rules at passive crossings as there would rarely be a train. Some participants
stated that they would drive through passive crossings in rural or remote areas even if
a train was visible as they believed that it would be easy to judge the speed and
distance of the train.
Risk factors for being involved in a collision
A number of near-miss stories were told by younger regional participants which
were said to be at bad crossings where there were numerous occurrences. It was
unknown if these occurrences were just by local residents or by tourists as well. The
general perception among regional participants was that these crossings were due
purely to design issues and were not due to risk taking by either themselves or other
motorists. Many regional participants discussed visibility, angle of approach and
actual awareness of being at a level crossing as two key safety issues in regional
areas. Issues of the poor visibility of the train and driving through crossings at night
were also mentioned by many participants. At some crossings, it is almost impossible
to see if a train is coming. Many younger regional participants raised their concern
that there is often little lighting on the sides of trains which make them difficult to see
at night. Influencing factors to drive safely at level crossings for younger regional
participants also included past experiences of near-misses.
Near-misses were considered to be a wake-up call for some regional
participants as they believed they may have become complacent with certain
crossings as there is a low volume of train traffic through the crossing on a daily
basis. Having friends or family members as passengers was another issue raised by
the majority of participants. It was suggested that when family members are present
in their vehicle (such as parents) then they would always obey the rules at crossings
and not take risks. When friends were passengers in their vehicle, some participants
(particularly female participants) indicated that they would want their friends to think
162
I am a good driver. Male regional participants appeared less likely to be worried what
their friends thought, but equally concerned with female participants what their family
thought of their driving at level crossings.
Pressure from other drivers whilst waiting at a level crossing was raised by
many of the urban participants. The majority of these participants indicated that they
felt that other drivers forced them to drive through level crossings whilst the boom
was lifting and the flashing lights still activated. Some participants even suggested
that vehicles behind theirs would beep their horn to hurry them across the level
crossing. One participant suggested that he would just give them the finger to let the
other motorist know that he wasn’t moving. Peers as passengers were also seen as a
key element in driving safely at level crossings, with many urban participants stating
they would take more risks when driving by themselves. The risk of damaging their
car was a big influence on whether participants drove safely at crossings or not. Some
participants stated their car was very important to them and if they were involved in a
collision with a train (and not injured) then they would be distressed if their vehicle
was destroyed as they had ‘saved up’ their money to purchase it. Some younger
urban participants indicated that if they were involved in a collision with a train that it
would be a slow speed (because the train would slow down) and that only their
vehicle would be damaged. Not everyone dies if their car gets hit, do they?
Driving on ‘auto-pilot’ was stated by a small number of urban participants as
being common at crossings they are familiar with. Complacency was suggested to be
common whilst driving generally, but some participants stated that they were very
complacent at local crossings that they drive over regularly. Being familiar with a
level crossing was deemed by most participants as making no difference to the level
of risk, although some participants stated that there is less risk at familiar crossings as
they know the risks.
Knowledge and perceptions of laws and enforcement
All younger regional participants were aware of the different types of crossings
due to the sample area having a range of both passive and active crossings. Road
rules at level crossings were known by the majority of regional participants and in
contrast to their urban counterparts, this group discussed a greater awareness of safety
issues at level crossings. There was a general perception that there was little
enforcement by police a crossings and a very low likelihood of being fined.
163
Participants indicated that police have other matters to attend to and that booking
motorists disobeying the road rules at level crossings was low on the policing agenda
in regional areas. You never see them out there.
The majority of younger drivers in the urban group stated that they were not
familiar with the type of level crossings outside of the metropolitan area as they rarely
drove in rural or remote areas. Some metropolitan participants even indicated that
they did not know that there were passive crossings with only a Stop or Give-way
sign. Among the urban group, there was generally a very low level of knowledge
about yellow hatching road markings that denote ‘keep clear’. Only a few
participants could correctly explain what such road markings mean. Many of the
urban participants indicated that they had never seen a crossing with such markings in
the metropolitan area. I’ve never even seen them, was one comment from a younger
urban participant. Knowledge about penalties and fines for breaches at crossings was
also quite poor in the urban group, with the general perception among these
participants that they have a very small chance of being caught for disobeying road
rules at crossing due to lack of police enforcement. Many of the urban participants
commented that they weren’t even aware that they could be fined for driving through
activated flashing lights or for queuing across a level crossing.
Usefulness of education and publicity
Nearly all of the regional participants thought that mass media campaigns were
the most effective method in changing their and others behaviour, with consequences
of a collision being an important message to many in this group. Most of the younger
regional participants reported that they listen to the radio when driving, and that using
radio as a delivery method would reach many younger drivers. Some participants
stated that they thought shock television campaigns were the best for young drivers
while other participants indicated that they would just tune out to shock style
campaigns. A suggestion for an education video for schools was also raised by many
of the regional participants. Some participants also mentioned including increasing
awareness about level crossing safety in the licensing process and when young drivers
learn to drive (such as through driving schools). One participant indicated that she had
never seen any questions about level crossings on her licence exam. Additionally, the
majority of participants in the regional group believed that it is also important to
continue to be made aware of risks at crossings, not just at the initial licensing.
164
Nearly all of the younger urban participants thought that mass media campaigns
were effective in changing their and others behaviour. A couple of female
participants stated that they would change the channel if such shock-style campaigns
came on the television and that such campaigns would only be effective for some
young drivers. Television commercials that clearly showed consequences and that
they could relate to on a personal level were seen as the most effective tool for
communicating to most members of the urban group. However, many participants
also stated that radio was also an effective tool as they frequently listen to the radio
whilst driving. Additionally, many members in the urban group indicated that they
need to be ‘given the information’ about the dangers of driving rather than having to
seek the information. Creating an awareness of the risks at level crossings was
thought by many as important in level crossing education. This included targeting
risk at passive crossings for urban drivers that travel to rural and remote areas (such as
going on a holiday). Some younger urban participants suggested providing
information such as how difficult it is to judge the speed of a train and how difficult it
is for a train driver to avoid a collision at a crossing. One participant commented that
approaching trains are like an illusion…all of a sudden they are there.
Differences between regional and urban group responses
Interestingly for the younger driver group, there appeared to be more
similarities in their responses than differences. The most notable difference between
these two groups was knowledge about different types of protection systems. Like the
older urban group, the younger urban group also had a poor level of knowledge about
passively protected crossings (i.e. Stop or Give-way signs only) outside of the
metropolitan area. Overall, there was very little difference between regional and
urban group responses in relation to perception of risk, reported driving behaviour,
risk factors for being involved in a collision with a train, perceptions of laws and
enforcement and the usefulness of education and publicity campaigns.
Both groups believe that they are at low risk of being involved in a collision
with a train as they believe that they have the skills and abilities to avoid a collision.
Interestingly though, some participants in both groups indicated that they find it
difficult to judge the distance a train is from a crossing which appears to oppose such
views of being able to avoid a collision. Additionally, there was very little difference
in reported driving behaviour, with many participants from regional and urban areas
165
stating that they frequently took risks at level crossings. Impatience for waiting at
crossings was evident in both of these groups, with regional drivers being frustrated
with the time it takes for freight trains to pass while urban drivers were frustrated with
waiting for protection systems (i.e. boom gates or flashing lights) to turn off. Like the
inconsistency in both group’s perceived level of risk at level crossings, there appeared
to be a paradox in self-reported driving behaviour. Most participants in both groups
stated that they frequently take risks at crossings (such as driving through flashing
lights or when a boom gate is descending), however the majority of participants
mentioned that other motorists caused them to take risks at crossings.
5.3.3.4 Heavy vehicle drivers
All participants in the truck driver group were male (n=26) reflecting the gender
profile of this workforce. Each driver took part in a semi-structured interview at a
major Brisbane Truckstop. Postcodes of drivers’ home addresses indicated that
participants were from a range of areas within the states of South Australia, Victoria,
New South Wales and Queensland. The mean age of the group was 46.3 years (range
26-68 years) and the mean years licensed as a driver was 29.2 years (range 8-54
years). The mean number of years driving trucks was 23.3 years (range 4-54 years).
Fifteen stated they were married / in a de facto relationship (57.7%) and the majority
were born in Australia (n=25, 96.2%). There were high levels of hours driven per
week reported, with 21 participants reporting driving more than 60 hours per week
(mean = 80.8 hrs per week).
When asked to estimate what percentage of the level crossings they encountered
were actively controlled, almost half the participants (n=12, 54.5%) stated ‘none’
were actively controlled, 22.7% (n=5) estimated that up to 40% were active and
22.6% (n=5) estimated 41-100% were active crossings. When asked specifically
about crossings that have flashing lights only and no boom barrier, 42.3% of
participants (n=11) estimated that 61-80% of the crossings were of this type. Twenty
three percent (n=6) of drivers reported being involved in a road crash during the past
five years, with all of those drivers reporting they were at fault in the collision. Traffic
offences history is illustrated in Table 10.
166
Table 10: Data collected from truck drivers during formative research phase Offence type n % Driving offences 24 92.3 Speeding offences 16 61.5 Log book offences 12 46.2 Other offences 5 19.1
Perception of risk
Level crossing safety was low on the scale of important road safety issues for
most of the truck driver participants, with some stating fatigue, drugs and speeding
were the key road safety problem areas. Other motorists were largely seen by
participants as causing many of the safety risks at level crossings. Although other
motorists were mentioned by most participants as increasing their risk of being
involved in a collision, very few participants elaborated further about how other
motorists place trucks at risk. Those participants that indicated there could be some
chance that they would be involved in a collision with a train specified design issues
and limitations of their truck as being paramount in increasing their risk.
On the other hand, participants that held the perception that there would be very
little if at all no chance of being involved in a collision with a train made statements
such as you don’t mess with trains, I always obey the rules at crossings and it would
be too costly. Interestingly, older truck drivers (i.e. over 55 years) appeared to hold
the perception that they could be involved in a collision compared to the younger
participants interviewed. The reasons for this may possibly be related to older truck
drivers being aware of more collisions (through both media and fellow colleagues)
than their younger counterparts. As such, they may no longer believe that they are
invincible or they may not be as confident driving a truck as they once were.
Many participants were aware of a colleague being involved in a heavy vehicle-
train collision, with some participants stating they would not divulge such
information. Some participants mentioned that the media played a role in ensuring
that any heavy vehicle-train collisions are viewed as both dramatic and potentially
catastrophic, even when the truck driver was not entirely to blame for the collision.
Self-reported and intended driving behaviour
Most participants stated that they always try to drive safely at level crossings,
with a handful of participants stating that they sometimes take risks. One participant
167
suggested that if there is no train near the crossing, then I might take a risk, while
another participant stated that out in the middle of nowhere, you can safely get across
without stopping. Those participants that admitted to taking risks at level crossings
said they would often try to ‘beat the train’ if they thought they would be waiting any
length of time at the crossing (e.g. for a freight train). Waiting at level crossings for
long trains to pass (such as a freight train with 40-50 wagons) was seen as frustrating
for some participants as it would ‘break their journey’. Sometimes there are 40
wagons on the backs of the trains. Many of these risk taking drivers held the
perception that it was more difficult to brake to stop at level crossings than it was to
just put your foot down.
Risk factors for being involved in a collision
There were a range of potential risk factors for driving a truck at level crossings.
Fatigue was listed by many participants as being one possible factor increasing the
likelihood of a collision. There appeared to be a general belief that within the
trucking industry there are a sub-set of truck drivers that ignore the warning signs of
fatigue which has an over-arching effect on the safety of driving (for both the truck
driver and other motorists). Some participants stated that it is part-and-parcel of
driving trucks, but that it can be managed either through having quality sleep, taking
breaks whilst driving or working for companies that abide by fatigue management
laws. There appeared to be differing views about the effectiveness of such fatigue
management laws, with one driver suggesting that it is far too limiting while another
saying that it was about time that something was done about drivers work ridiculous
hours. A few of the older participants (over 55 years) stated that their younger
counterparts consume either illicit or licit substances in order to alleviate work-related
fatigue. Many of the younger participants would not comment on such issues,
although one young participant suggested that some drivers do what it takes to stay
awake at the wheel.
Complacency was also listed as a possible factor increasing the likelihood of a
collision. Complacency was mentioned to be linked with driving outside the
metropolitan area such as in rural or remote areas. Several participants mentioned
that some truck drivers may hold the belief that I’ve never seen a train here yet at
certain level crossings with small train volumes, while other participants mentioned
that some truck drivers may hold the belief that a train will slow down to avoid a
168
collision with a truck. There was a general consensus among participants that
complacency is not just within the trucking industry but extends to the general
motorist. Like complacency, distraction was seen by many participants as being a
risk factor for driving trucks generally and at level crossings. Distraction was seen by
many of the participants as being related to in-vehicle distraction such as wireless
communication devices (such as map and other complex visual displays), talking on a
mobile phone or eating/drinking. These tasks were thought to impair a truck driver’s
visual search patterns, reaction times and decision-making processes, causing them to
drive unsafely on the road generally or at level crossings.
Judgment errors, such as a truck not clearing a crossing (when the trailer of the
truck extends onto or over the rail tracks and is known as ‘short-stacking), were
believed by many to be a problem that will just get bigger with the increase in vehicle
lengths in Australia. Some drivers indicated that they sometimes have difficulty in
knowing if their trailer has completely cleared a crossing, with a few stating that this
issue can make level crossings hazardous for trucks. Judgment errors appeared to be
linked to how familiar a truck driver is with a particular level crossing. Being
unfamiliar with a particular level crossing was another factor raised by many
participants as being a possible factor in causing a collision. Some participants
suggested that if they hadn’t crossed a level crossing before then they would be
uncertain if their trailer would over-hang the crossing and this would pose a safety
risk. No participant hinted at the stress that a train driver may feel if they are
approaching a level crossing when a truck is over-hanging the tracks.
Blinding sun or faulty protection systems were mentioned as being key reasons
why truck drivers couldn’t be blamed for causing the collision. These two factors
were mentioned by many of the interview participants as being problematic for
driving at level crossings. Some participants recalled hearing of collisions that
involved a truck driver not being able to see activated flashing lights and suggested
that this is a common problem due to the height of the cabin in relation to the height
of the warning systems. Other pressing design and engineering issues for participants
included angle of approach, visibility restrictions, road surfaces, and inadequate
stopping distances. Of these issues, the angle of approach and visibility were seen as
the most problematic for the majority of drivers. Participants indicated that crossings
that have an approach angle that is oblique (i.e. an angle, such as an acute or obtuse
169
angle, that is not a right angle or a multiple of a right angle) would be sub-optimal for
safely negotiating their truck. Dog-leg approaches are the worst.
Knowledge and perceptions of laws and enforcement
This truck driver group displayed a very high level of knowledge of the road
rules and types of level crossings that exist. Most participants could recall the location
of crossings that they regularly drove through on their designated routes. All
participants knew the road rules at level crossings and with statements such as trains
always have right of way and we respect trains. Many of the participants talked about
trains being a larger mass than their heavy vehicle and the fact that it would be messy
if they collided with a train. The majority of participants indicated that there is a very
low level of enforcement by police at crossings, however all believe police
enforcement would be ineffective as drivers would inform other drivers on the two-
way radio (UHF). One participant suggested that word would spread quickly. Many
participants indicated seeing some enforcement at level crossings during their careers,
however they all believed that unless there are cameras installed to observe for
violation of road rules then enforcement is a waste of time and money.
Improving safety for trucks
In terms of improving safety for trucks at level crossings, advance warning
systems were seen as a key by all participants to reducing the likelihood of a collision.
Many participants also mentioned that trains need to be better lit up at night as they
are difficult to see, particularly when a truck is driving parallel to a train as it is
difficult to see the headlights, with one participant suggesting that it is impossible to
see trains at night. Most participants indicated that it is also impossible to hear a train
approaching a level crossing as the engine underneath a truck cabin is so noisy that
you can’t hear yourself think sometimes. As such, it was suggested that warning bells
and alarms should be made louder at crossings with large volumes of heavy vehicle
traffic.
Usefulness of education and publicity
The majority of participants had never experienced any form of level crossing
safety education and comments indicated that both engineering and education needs
should be addressed for truck driver safety:
170
Fix physical road issues first and follow with education – education is pointless
if you can’t perform it.
Make people safe. You can make the crossing as safe as you like but its people
that do the unsafe things, make the people safe.
While a number of participants made positive or insightful comments about
educating truck drivers about the dangers of level crossings, a small number felt that
education was not the answer. Most participants indicated that financial efforts
shouldn’t be targeted towards their driving behaviour at crossings, but rather that of
other motorists. Additionally, some participants suggested that educating train
drivers was also an important tool for improving their own safety at crossings. Some
participants held the belief that train drivers should be made more aware of the
difficulties that trucks face at level crossings (such as braking and accelerating times)
as they can slow down if they want to.
Those participants that held the perception that education was needed to inform
and educate truck drivers about the safety of issues of driving trucks at level
crossings, raised a variety of delivery methods for reaching truck drivers. One
method that was considered acceptable to truck participants was an article in trucking
magazines (e.g. ‘Big Rig’, ‘Truck and Life’ or ‘Owner Driver’). Most participants
stated that they read trucking magazines during breaks and thought they were a
credible source of information. Articles that focused on ‘real-life experiences’ were
considered the most interesting to read. Existing training sessions with trucking
companies was also suggested by many participants as being an appropriate method
for delivery of education. Other ideas included radio safety advertisements, billboards
near level crossings or posters at truckstops.
5.3.4 Discussion
As stated previously, information obtained from each of the three road user
groups was critical in assisting, guiding, and identifying priority areas for message
and material development, as well as the development of data collection measures
(i.e. questionnaires). The reportage of the results of both interview and focus groups
was broken down into key themes that were generated by the data. These themes
171
provided the framework for the categorisation of the responses, with participants’ own
words being identified in the text by the use of italics.
The results of the qualitative formative research from each of the three road user
groups revealed some interesting differences between the three groups. For the
younger drivers however, there were more similarities than differences between
responses from regional and urban participants. Younger drivers demonstrated low
risk perception of the consequences to unsafe driving behaviour at level crossings,
with risk taking being reported at high levels for this group. The regional group
reported a high acceptance of risk taking behaviours such as trying to ‘beat the train’,
while participants in the urban group reported driving through activated protection
systems (either before or after a train had passed through the level crossing). Both
groups however held the belief that they could such risks were acceptable as they
consider that they have the skills and abilities to safely do so. As such risk perception
and risk acceptance are the two key issues that were highlighted in researching this
group. The most acceptable method of delivering educational interventions for this
group appears to be through radio or television. A scenario involving risking their
friends’ lives when driving unsafely was a common theme for developing messages.
For the older participants, most displayed high levels of knowledge and reported
low risk taking behaviour, with age-related factors being acknowledged by the group
as being risk factors for their involvement in a collision. However, this group
indicated that they employed compensatory and protective measures to reduce or
control their level of risk (e.g. driving during the day, going slower, being more alert,
and avoiding crossings they felt were dangerous). According to the participants in the
older group, educational strategies would best be focused on reminder or re-affirming
messages as they believe they always obey crossing rules and try to drive as safely as
possible. Radio messages or public talks at social groups appeared to be the most
acceptable method of delivery for educational campaigns for the majority of
participants in the older group.
Heavy vehicle drivers indicated a high level of knowledge of safety issues, with
engineering and design being a major factor contributing to near-misses or the risk of
being involved in a collision at a level crossing. Risk taking was acknowledged as
minority behaviour in the industry, with time pressures, complacency, monotony,
inattention and fatigue noted to be risk factors. There were mixed opinions about
educational interventions targeted towards heavy vehicle drivers, with most drivers
172
indicating that the most acceptable methods would be in sharing of personal
experiences (through trucking magazines) and awareness raising (through company
training sessions).
A comparison of the three road user groups indicates that only younger drivers
admit to violating the road rules at level crossings. Both regional and urban younger
drivers admit to past risk taking behaviours at crossings as well as holding intentions
to take risks in the future. Older drivers typically reported that their skills and
abilities to drive safely at level crossings were sometimes reduced due to degeneration
of their vision, hearing and reflexes. Many older drivers suggested that they
‘compensate’ for these difficulties by driving at familiar crossings during times that
cause them least stress (i.e. during the day). Older drivers also generally believe that it
is other motorists that take risks at crossings and pressure them to drive unsafely.
Heavy vehicle drivers appear to be the most familiar group with the different type of
protection systems, which is not surprising given their exposure to driving in both
urban and regional areas. However, this group held the strongest beliefs that
engineering and design of crossings caused them to potentially drive unsafely due to
limitations of their vehicle (e.g. visibility restrictions, ability to quickly clear a
crossing). Although many truck drivers held strong beliefs regarding their plight
with warning and protection systems at level crossings, they also acknowledged that
they may become complacent when driving due to time pressures, fatigue and
monotony.
These findings highlight the continual need to acknowledge the distinction
between performance (the abilities of drivers to perceive and react to circumstances in
an appropriate and timely manner) and behaviour (what drivers are able to do,
including the perception and acceptance of risk, as well as peer pressure). With
research suggesting that behaviour change not being among the outcomes that can be
accomplished by educational interventions alone (Henderson, 1991), level crossing
interventions may achieve an increase in the awareness of the safety problem or
behaviour, assist in the formation of beliefs (especially when not held formally), and
contribute to the ‘road safety culture’ in the community (Henderson, 1991, Di Pietro,
2003).
173
5.4 QUANTITATIVE RESEARCH WITH TRAIN DRIVERS AND EXPERTS
5.4.1 Objectives
Although the qualitative formative research assisted in understanding the
meanings that the target groups assign to level crossing safety as well as exposing the
psychological processes underlying unsafe behaviours, quantitative research was
required to help interpret these qualitative findings as well as cross-validate and
quantify at risk behaviours.
5.4.2 Method
5.4.2.1 Ethical clearance
Ethical clearance for data collection for this feedback phase was gained from
the Queensland University of Technology Human Research Ethics Committee (QUT
Ref. No. 3550H).
5.4.2.2 Sampling
A non-random sampling technique (convenience sampling) was also used for
this quantitative formative research. Although non-random samples are not able to be
analysed for significance, this type of non-probability sampling was the most
appropriate and practical within the budget and resources of the project.
A large rail organisation in Queensland was the subject pool for train drivers.
These drivers were identified through the organisation’s Train Crew Managers in both
metropolitan and regional areas within Queensland. Train Crew Managers were each
sent 30 questionnaires to distribute to their train drivers (i.e. 60 questionnaires sent
out in total). A list of experts in level crossing safety in Australia was compiled in
consultation with the Australasian Railway Association and state rail authorities. This
list was used for the modified Delphi technique questionnaires in Study 1. Both of
these groups completed questionnaires to ascertain any differences between beliefs of
train drivers and experts in the field.
5.4.2.3 Procedure
Information sheets were provided with each survey questionnaire sent to both
train drivers and experts. Participants were informed that participation was purely
174
voluntary and that their responses would remain anonymous. Consent was implied
with return of the completed questionnaire. The questionnaire took approximately 10
minutes to complete. Participants were provided with a reply-paid envelope in which
to return their completed questionnaire.
For the train drivers, supervisors in both the regional and urban area were
mailed questionnaires each during February/March 2006 to distribute to train drivers.
All questionnaires were returned within a few weeks from being mailed. For the
experts in the field, each expert was mailed a survey questionnaire during January
2006. A reminder letter was sent a few weeks later to obtain a maximum response
rate.
5.4.2.4 Materials
A 19-item questionnaire using a Likert scale (1=low risk; 5=very high risk) was
developed for this quantitative formative research. The survey questionnaire sent to
both train drivers and experts are provided in Appendix 4. This questionnaire was
developed from results obtained from the qualitative formative research conducted
with each of the three road user groups. Themes that emerged from this qualitative
research (i.e. high risk behaviours of drivers) were selected by a panel of 3 road safety
experts (separate to those experts participating in this survey). Reliability analysis
was undertaken to provide information about the relationships between individual
items in the 19-item instrument. Reliability is considered to be the correlation of an
item, scale or instrument with a hypothetical one which truly measures what it is
supposed to (McGraw and Wong, 1996). The set of 19 items shows good internal
consistency with this sample, recording a Cronbach alpha value of .860.
5.4.2.5 Data analysis
The data collected was analysed using the Statistical Package for Social
Sciences (SPSS) Version 15.0. Given the nature of the survey questionnaire, there
was minimal missing data, with cases not being excluded from analyses if they had
missing data due to the minimal likelihood of bias.
175
5.4.3 Results
A total of 47 train drivers from both regional (n=22, 73.3%) and urban (n=25,
83.3%) areas agreed to participate in this feedback survey, while thirty six (36) survey
questionnaires were sent to experts in the field, with 32 completing and returning the
questionnaire (88.8%). The response rates for both groups were above average for
both groups. No demographic data was collected from the survey questionnaire sent
to either train drivers or the expert panel. Both of these groups had been surveyed
previously on two occasions using the modified Delphi Technique, and it was not
deemed relevant in this formative quantitative research phase.
A number of results are presented in this section. Firstly, an examination of the
differences between urban and regional train driver risk ratings. An independent t-test
was used to assess the differences in overall mean scores as well as individual items.
Additionally, these risk ratings were ranked (by mean order) from highest (mean
score greater than 4.50) to moderate (mean score of 3.50 or less) risk. Secondly, an
examination of the differences between train drivers (aggregate) and expert’s risk
ratings. An independent t-test was also used to assess the differences in overall mean
scores as well as individual items. Additionally, these risk ratings were ranked (by
mean order) from highest (mean score greater than 4.50) to moderate (mean score of
3.50 or less) risk. Each of these comparisons is displayed in tables.
Table 11 presents the findings from the aggregate data from both urban and
regional train drivers. As can be seen in Table 11 below, aggregate ratings of risk
proscribed by the train drivers leaned toward the more dangerous end of the spectrum.
To this end, 10 of the 19 acts were regarded as belonging to the highest echelon of
dangerous behaviours at level crossings, polling an average risk rating of between
4.51 and 5.00. Not surprisingly, at the head of this list was the act of trying to beat
the train across the crossing, where the driver knowingly gambles on their ability to
safely clear a crossing before an approaching train crosses. Also in the highest risk
category were behaviours that implied crossing in poor visibility situations, crossing
with undue haste (overtaking, bypassing boom gates, crossing in-front of a visible
train), crossing without due attention (not looking for a second train, queuing, not
noticing passive crossings) and ignoring warning devices (not stopping for passive
crossings, driving through an activated active crossing). The next seven behaviours
were regarded by the drivers as high risk, and polled an average risk rating of between
176
4.50 and 3.51. These behaviours again demonstrated the common ground of acts of
undue haste or negligence, but seemingly also implied a level of safety consideration.
For example, included behaviours such as scanning on approach without stopping and
crossing when the train is distant are obviously dangerous, but demonstrate at least
some consideration of risk by the driver. The two remaining behaviours, crossing
before boom gates began to descend or had fully ascended, were regarded as
presenting only moderate levels of risk. This is likely because, though these
behaviours are against the law and hold the potential for harm, they do show some
deference to the active protections.
177
Table 11: Level of aggregate train driver assigned risk to motorist behaviours Item
Aggregate Class Rating
Aggregate Mean
Aggregate S.D.
Try to beat the train across the crossing Highest 4.91 0.28 Driving around the boom gates Highest 4.81 0.68 Drive in front of train when it is 'close' to the crossing Highest 4.77 0.52 Going thru the crossing as soon as one train has passed Highest 4.74 0.57 Queuing up over a congested crossing Highest 4.72 0.71 Not looking at passive crossings Highest 4.68 0.51 Go thru flashing lights (at crossings with flashing lights only) when train visible Highest 4.64 0.60 Going across the crossing when unable to see if there is a train coming (poor visibility) Highest 4.62 0.61 Not stopping at all at passive crossings Highest 4.57 0.54 Overtaking cars that are stopped at the crossing Highest 4.55 0.97 Following the car in front across the crossing without looking High 4.43 0.71 Trying to get through the crossing before the boom gates come down (moving) High 4.30 0.93 Speeding on approach to crossings High 4.21 0.83 Going through flashing lights High 4.11 0.73 Going at passive crossings when the train is visible but 'far away' High 3.79 0.88 Looking or scanning on approach and then not stopping or slowing if no train seen High 3.68 0.86 Slowing and rolling thru STOP signed crossings High 3.62 1.09 Going thru flashing lights before boom gates start to go down Moderate 3.43 1.35 Going thru flashing lights before boom gates start to come up Moderate 2.85 1.30
178
The findings from the comparison of risk ratings between urban and regional
train drivers found some interesting trends. When Tables 12 and 13 below are
compared, the similarities and differences of the drivers’ ratings are drawn into focus.
At first glance the ratings appear fairly consistent between the two groups. The act of
consciously trying to beat the train across the tracks again draws most severe ratings,
and again the behaviours of crossing before boom gates begin to descend or have fully
ascended again rate as the least dangerous. Similar also is the risk ratings skew
toward rating the behaviours as generally dangerous, and the distribution of numbers
of behaviours between the highest, high and moderate risk groups. An independent t-
test was also conducted to compare the risk ratings of train drivers. An overall mean
score was computed for the 19 items. Table 14 presents the statistics for each item.
The mean score for regional train drivers was 4.28 (n=22) while the mean score for
the urban group was 4.29 (n=25). There was no significant difference between these
mean scores [t(45)= -.121, p=.904,ns]. Only one item ‘go through flashing lights (at
crossings with flashing lights only) when a train is visible’ was found to have a p-
value of 0.05.
179
Table 12: Level of regional train driver assigned risk to motorist behaviours Item
Regional Class Rating
Regional Mean
Regional S.D.
Try to beat the train across the crossing Highest 4.86 0.35 Going thru the crossing as soon as one train has passed Highest 4.77 0.43 Driving around the boom gates Highest 4.68 0.94 Drive in front of train when it is 'close' to the crossing Highest 4.68 0.65 Going across the crossing when unable to see if there is a train coming (poor visibility) Highest 4.64 0.66 Queuing up over a congested crossing Highest 4.64 0.95 Following the car in front across the crossing without looking Highest 4.59 0.67 Not looking at passive crossings Highest 4.55 0.60 Not stopping at all at passive crossings High 4.50 0.60 Trying to get through the crossing before the boom gates come down (moving) High 4.50 1.01 Go thru flashing lights (at crossings with flashing lights only) when train visible High 4.45 0.67 Overtaking cars that are stopped at the crossing High 4.45 1.18 Speeding on approach to crossings High 4.32 0.78 Going through flashing lights High 4.05 0.78 Looking or scanning on approach and then not stopping or slowing if no train seen High 3.73 0.88 Going at passive crossings when the train is visible but 'far away' High 3.73 0.88 Slowing and rolling thru STOP signed crossings High 3.59 1.18 Going thru flashing lights before boom gates start to go down Moderate 3.50 1.14 Going thru flashing lights before boom gates start to come up Moderate 3.05 1.25
180
Table 13: Level of urban train driver assigned risk to motorist behaviours Item
Urban Class Rating
Urban Mean
Urban S.D.
Try to beat the train across the crossing Highest 4.96 0.20 Driving around the boom gates Highest 4.92 0.28 Drive in front of train when it is 'close' to the crossing Highest 4.84 0.37 Go thru flashing lights (at crossings with flashing lights only) when train visible Highest 4.80 0.50 Queuing up over a congested crossing Highest 4.80 0.41 Not looking at passive crossings Highest 4.80 0.41 Going thru the crossing as soon as one train has passed Highest 4.72 0.68 Not stopping at all at passive crossings Highest 4.64 0.49 Overtaking cars that are stopped at the crossing Highest 4.64 0.76 Going across the crossing when unable to see if there is a train coming (poor visibility) Highest 4.60 0.58 Following the car in front across the crossing without looking High 4.28 0.73 Going through flashing lights High 4.16 0.69 Speeding on approach to crossings High 4.12 0.88 Trying to get through the crossing before the boom gates come down (moving) High 4.12 0.83 Going at passive crossings when the train is visible but 'far away' High 3.84 0.90 Slowing and rolling thru STOP signed crossings High 3.64 1.04 Looking or scanning on approach and then not stopping or slowing if no train seen High 3.64 0.86 Going thru flashing lights before boom gates start to go down Moderate 3.36 1.52 Going thru flashing lights before boom gates start to come up Moderate 2.68 1.34
181
Table 14: Comparison of means between urban and regional train drivers
Mean Item Urban Regional t df Sig Driving through the crossing when lights flashing before boom gates go down 3.36 3.50 .352 45 .726 Trying to get through the crossing before the boom gates come down (moving) 4.12 4.50 1.412 45 .165 Going through flashing lights before boom gates start to come up 2.68 3.05 .959 45 .342 Slowing and rolling thru STOP signed crossings 3.64 3.59 -.152 45 .880 Going thru the crossing as soon as one train has passed 4.72 4.77 .313 45 .755 Not stopping at all at passive crossings 4.64 4.50 -.882 45 .382 Not looking at passive crossings 4.80 4.55 -.1726 45 .091 Going at passive crossings when the train is visible but 'far away' 3.84 3.73 -.433 45 .667 Drive in front of a train when it is 'close' to the crossing 4.84 4.68 -1.042 45 .303 Queuing up over a congested crossing 4.80 4.64 -.781 45 .439 Driving around the boom gates 4.92 4.68 -1.204 45 .235 Trying to beat the train across the crossing 4.96 4.86 -1.173 45 .247 Speeding on approach to crossings 4.12 4.32 .811 45 .421 Overtaking cars that are stopped at the crossing 4.64 4.45 -.647 45 .521 Looking or scanning on approach and then not stopping or slowing if no train seen 3.64 3.73 .343 45 .733 Going through flashing lights 4.16 4.05 -.533 45 .597 Go through flashing lights (at crossings with flashing lights only) when a train is visible 4.80 4.45 -2.017 45 .050 Following the car in front across the crossing without looking 4.28 4.59 1.509 45 .138 Going across the crossing when unable to see if there is a train coming (poor visibility) 4.60 4.64 .202 45 .841 Global Mean Score 4.29 4.28 -.121 45 .904
182
Table 15 illustrates the assigned risk levels by the panel of experts. Experts
ranked ‘driving around boom gates’ as the highest risk behaviour by motorists, with a
mean of 4.91 (n=32). The lowest ranked item was ‘going through flashing lights
before the boom gates start to go down’, with a mean of 2.63 (n=31). Additionally, an
independent t-test was conducted to compare overall mean scores between train
drivers and experts. This mean score was computed for each case to provide an
overall mean of the 19-items. Statistics for each item is displayed in Table 16. The
mean score for the expert panel was 4.06 (n=32) while the mean score for the train
drivers was 4.29 (n=47). There was no significant difference between these mean
scores [t(77)=-2.334, p=.998, ns]. Only one item ‘overtaking cars that are stopped at
the crossing’ was found to have a p-value of less than 0.05. The p-value for this item
was .001.
183
Table 15: Level of expert assigned risk to motorist behaviours Item Class Rating Mean S.D. Driving around the boom gates Highest 4.91 .296 Try to beat the train across the crossing Highest 4.90 .396 Overtaking cars that are stopped at the crossing Highest 4.88 .421 Drive in front of train when it is 'close' to the crossing Highest 4.84 .374 Queuing up over a congested crossing Highest 4.78 .420 Going thru the crossing as soon as one train has passed Highest 4.72 .634 Not looking at passive crossings Highest 4.66 .602 Go thru flashing lights (at crossings with flashing lights only) when train visible Highest 4.63 .660 Going across the crossing when unable to see if there is a train coming (poor visibility) High 4.34 .701 Following the car in front across the crossing without looking High 4.22 .608 Not stopping at all at passive crossings High 4.13 .833 Going through flashing lights High 4.00 .816 Trying to get through the crossing before the boom gates come down (moving) High 3.94 1.014 Speeding on approach to crossings High 3.90 .870 Going at passive crossings when the train is visible but 'far away' Moderate 3.40 .932 Looking or scanning on approach and then not stopping or slowing if no train seen Moderate 2.88 1.070 Slowing and rolling thru STOP signed crossings Moderate 2.75 1.218 Going thru flashing lights before boom gates start to come up Moderate 2.65 1.082 Going thru flashing lights before boom gates start to go down Moderate 2.63 1.185
184
Table 16: Comparison of means between train drivers and experts
Mean Item Experts Train
Drivers t df Sig Driving through the crossing when lights flashing before boom gates go down 2.63 3.43 -2.720 77 .180 Trying to get through the crossing before the boom gates come down (moving) 3.94 4.30 -1.629 77 .821 Going through flashing lights before boom gates start to come up 2.65 2.85 -.730 76 .374 Slowing and rolling thru STOP signed crossings 2.75 3.62 -3.301 77 .312 Going thru the crossing as soon as one train has passed 4.72 4.74 -.190 77 .780 Not stopping at all at passive crossings 4.13 4.57 -2.909 77 .062 Not looking at passive crossings 4.66 4.68 -.195 77 .481 Going at passive crossings when the train is visible but 'far away' 3.40 3.79 -1.836 75 .532 Drive in front of a train when it is 'close' to the crossing 4.84 4.77 .672 76 .145 Queuing up over a congested crossing 4.78 4.72 .412 77 .312 Driving around the boom gates 4.91 4.81 .764 77 .120 Trying to beat the train across the crossing 4.90 4.91 -.152 76 .703 Speeding on approach to crossings 3.90 4.21 -1.579 76 .890 Overtaking cars that are stopped at the crossing 4.88 4.55 1.758 77 .001 Looking or scanning on approach and then not stopping or slowing if no train seen 2.88 3.68 -3.695 77 .518 Going through flashing lights 4.00 4.11 -.601 76 .929 Go through flashing lights (at crossings with flashing lights only) when a train is visible 4.63 4.64 -.092 77 .925 Following the car in front across the crossing without looking 4.22 4.43 -1.339 77 .055 Going across the crossing when unable to see if there is a train coming (poor visibility) 4.34 4.62 -1.840 77 .192 Global Mean Score 4.06 4.29 -2.334 77 .998
185
5.4.4 Discussion
Data from both urban and regional train drivers provided an insightful comparison
of assigned risk levels for a list of commonly observed level crossing behaviours by
motorists. Rated as most dangerous was the act of trying to beat the train across the
crossing, a behavioural description that implies conscious risk taking on the part of the
motorist. Similar themes were repeated among the other upper echelon of dangerous
behaviours, specifically crossing with undue haste, crossing in poor visibility situations
and ignoring warning devices. These behaviours can be categorised as violations on the
part of the motorists, acts that imply a willful disregard of safe crossing protocols by the
motorists, as well as indifference toward the risks involved in their behaviours. The
remaining behaviours in this upper echelon of risk were indicative of negligent lapses,
where motorists are crossing without due attention to the conditions in front of them.
While these behaviours do not involve deliberate violation on the part of the motorist,
they remain deserving of examination as dangerous behaviours for the reason that they
are indicative of a passive approach to risk assessment while driving. Hence, at an
aggregate level, it is apparent that train drivers consider motorists’ deliberate protocol
violations and negligently lax approach to hazard detection as the predominant causes of
danger at level crossings (Lawton, 1997b, Caird, 2002). Mirroring the aggregate level
analysis, the assigned risk levels show noteworthy consistency between the urban and
regional groups. With only one exception, each of the level crossing behaviours retain
the same risk categorisation. Where discrepancies are observed, they appear on the
surface to be mostly minor differences in the rank order which the items appear. This is
notable in that it indicates that the both sets of train drivers, presumably drawing on their
own experiences of either urban or regional locales, still share a similar view of where the
danger lies at levels crossings.
The comparison between the aggregate assigned risk levels by train drivers and
those assigned by experts also revealed some interesting results. Both the panel of
experts and train drivers ranked the items ‘driving around the boom gates’ and ‘trying to
beat the train across the crossing’ as being the highest risk taking behaviours. Experts
also ranked ‘overtaking cars stopped at the crossing’ as being very high risk (4.88), while
186
train drivers ranked ‘drive in front of a train when it is close to the crossing’ (4.77) as
being a very high risk taking behaviour. Overall, experts were observed to rank the
behaviours lower (4.06) than train drivers (4.29). This may be due to the nature of train
drivers observing a large number of near-misses throughout their careers and the
possibility that they experience great frustration with motorists that take risks at level
crossings (findings from the focus groups discussions with train drivers in both urban and
rural areas). Additionally, experts in the field may have inadvertently made a comparison
of risk levels at level crossings and risk levels whilst driving generally. This would no
doubt have produced lower mean scores than those reported by train drivers.
5.5 FRAMEWORK FOR INTERVENTION DEVELOPMENT
5.5.1 Overview
The planning, development and delivery of behavioural interventions to road user
groups can be applied using two different approaches. ‘Social Marketing’ techniques and
‘Intervention Mapping’ are two techniques that have been applied in the delivery of
behavioural interventions. However, although social marketing principles have been
widely used in public health and social psychology, its basis is not driven by theory. This
is a major limitation of this approach, and its use is therefore restricted in developing
behavioural interventions. However, for the purposes of comparison, the merits and
limitations of both techniques will be reviewed.
It must be noted though that as this thesis did not have a sole purpose of evaluating
the effectiveness of the behavioural interventions developed in Study Three, intervention
mapping was not used in a prescriptive manner. As such, the systematic investigation of
the merit, worth, or significance of the interventions was not examined. What
intervention mapping did provide though was the emphasis on two significant factors that
are required: (1) combining theory with evidence (including industry experts such as
those used in Study One of this thesis) and (2) recognising that individual behaviour is
influenced by the environments in which individuals live, their social networks,
187
organisations, communities and societies. A detailed comparison of intervention mapping
with the popular social marketing techniques is provided.
5.5.2 Social marketing
Social marketing is the use of commercial marketing principles to promote the
adoption of a behaviour that will improve the health or well-being of the target audience
or of society as a whole (Weinreich, 1999, Nutbeam, 2004). To be effective, road safety
programs must precisely specify its target audience, use customised methods to reach the
target audience and involve the target audience in developing the program (Nutbeam,
2004). Social marketing does not rely purely on educating the target audience about an
issue, but rather uses persuasive messages developed through formative research with
members of the target audience (Weinreich, 1999). This framework involves research and
re-evaluation at each stage to assess if the program is developing appropriately (Nutbeam,
2004, Weinreich, 1999). According to Weinreich (1999) this process consists of five
general stages:
• Planning;
• Message and materials development;
• Pre-testing;
• Implementation; and
• Evaluation and feedback.
This process is non-linear but rather of feedback and adjustment that may require
re-visiting past stages to make changes based on new information (Weinreich, 1999).
The first step, the planning phase, forms the foundation on which the process is built.
This phase must be based on understanding the problem being addressed, the targeted
audiences, and the environment in which the program will operate (Weinreich, 1999).
Formative research is used to analyse such factors and develop a workable strategy for
influence behaviour change. Step Two uses information learned in the formative planning
phase to design messages to be conveyed to the target audiences, as well as the materials
to carry such messages (Weinreich, 1999). The pre-testing phase (Step Three) involves
188
using a range of methods to test messages, materials and proposed strategies with the
target audience members to determine what works best to achieve objectives of the
program (Weinreich, 1999). Step Four involves introducing the program to the target
audiences (Weinreich, 1999). Preparation and monitoring the implementation is
paramount to ensuring success of the program. Step Five, the evaluation and feedback
phase, assesses the effects of the program as a whole as well as the individual elements of
the strategy (Weinreich, 1999). Evaluation occurs throughout the process of program
development, and feedback is used at each stage to improve the program (Weinreich,
1999).
Fishbein (2001) proposes that while there has been a growing recognition of the
value of theory-based behavioural interventions, it is seldom clear how theory is actually
used in the development (and evaluation) of such interventions. Intervention mapping is a
protocol for the development of theory-based behavioural intervention, by providing
guidelines and tools for the application of theory as well as the translation of theory in
actual program materials and activities (Bartholomew et al., 1998, Hoelscher et al.,
2002).
5.5.3 Intervention mapping
Intervention mapping also provides researchers and program planners with a
framework for effective decision-making at each step in intervention planning,
implementation and evaluation (Bartholomew et al., 2006). This concept is “based on the
importance of planning programs that are based on theory and evidence” (Bartholomew
et al., 2006, p8). The term ‘evidence’ refers to data from research studies as represented
in scientific literature as well as opinion and experience of community members and
industry experts (Bartholomew et al., 2006). As such, theoretical and empirical evidence
is brought together to meet a road safety need. The concept of intervention mapping
provides a detailed framework for this process to occur in a systematic and logistical
approach.
The primary concept of intervention mapping is that individual behaviour is
influenced by different causes at various environmental levels (such as environments in
189
which individuals live, including family, social networks, organisations, communities and
societies) (Kok et al., 2004, Bartholomew et al., 1998, Bartholomew et al., 2001).
Intervention mapping consists of five key steps: (1) creating a matrix of proximal
program objectives, (2) selecting theory-based intervention methods and practical
strategies, (3) designing and organising a program, (4) specifying adoption and
implementation plans, and (5) generating program evaluation plans (Bartholomew et al.,
1998). Intervention mapping has been used to develop intervention programs for a variety
of health behaviour programs including asthma management, sun protection, nutrition,
adolescent risk taking, cancer screening, HIV prevention and acute stroke therapy.
Although intervention mapping has not been found in the literature to have been applied
to road safety interventions, the merits of intervention mapping must not be overlooked.
The capacity of intervention mapping to provide road safety program planners with the
ability to integrate the “wealth of information, theories, ideas, and models to develop
interventions that are logical and appropriate in their foundations and are practical and
acceptable in their administration” is underestimated.
Intervention mapping provided valuable information to assist with the guiding and
identifying priority areas for message and material development in Study Three.
However, as this program of research did not have a sole purpose of evaluating the
effectiveness of the behavioural interventions, it was not used in a prescriptive manner.
Hence, the systematic investigation of the merit, worth, or significance of the
interventions was not examined. What intervention mapping did provide though was the
emphasis on two significant factors that are required: (1) combining theory with evidence
(including industry experts such as those used in Study One of this thesis) and (2)
recognising that individual behaviour is influenced by the environments in which
individuals live, their social networks, organisations, communities and societies. This
environmental change concept is important in its application to road safety, and
particularly important in attempting to improve level crossing safety in the heavy vehicle
industry.
190
5.6 SUMMARY
This chapter documented Study Two, the formative research undertaken as part of
the planning and development of interventions for each of the three road user groups, as
well as the development of data collection measures. As stated previously, intervention
mapping was used to assist in the planning and development of the interventions, but was
not used in a prescriptive manner. The next chapter will provide findings from Study
Three. Study Three involved the examination of the present context of motorist
behaviour at level crossings as well as the planning, development and measurement of
change associated with exposure to the interventions for each of the three road user
groups.
191
CHAPTER SIX: THE PRESENT CONTEXT OF MOTORIST BEHAVIOUR AT LEVEL CROSSINGS
6.1 Introduction …………………………………………………………… 192
6.2 Study aims and research questions ……………………………………. 192
6.3 Intervention development and implementation ……………………….. 194
6.4 Method ………………………………………………………………... 197
6.5 Results ………………………………………………………………… 215
6.6 Study limitations ……………………………………………………… 268
6.7 Summary ……………………………………………………………… 272
192
6.1 INTRODUCTION
This chapter documents the third study undertaken as part of the program of
research. There were three parts to this final study. The aim of Part One of this study was
to develop targeted interventions specific to each of the three road user groups in
accordance with Fishbein’s theoretical model (Integrated Model of Behaviour Change).
The development of interventions was originally seen as being outside of the scope of
this project, however, it became intertwined in questionnaire development and thus
deemed to be within the realms of the current mode of inquiry. The aim of Part Two was
to investigate the present context of unsafe driving behaviour at level crossings. This
included examining the personal, social and environmental factors contributing to unsafe
driving intention and driving behaviour at level crossings. The aim of Part Three was to
trial a pilot road safety radio advertisement using an intervention and control
methodology.
The Integrated Model of Behaviour Change (IM) developed by Martin Fishbein,
was useful in this exploratory investigation. Firstly, it provided a foundation on which to
build and develop each of the three intervention messages. Secondly, it assisted in
designing questionnaires based on the key variables of Fishbein’s model (i.e. intentions;
attitudes; norms; self-efficacy or perceived behavioural control; perceived risks; skills or
abilities; and environmental constraints). The design of these questionnaires was based on
the premise that intentions alone are not the singular determinant of behaviour, as
motorist behaviour at level crossings appears to be influenced by other factors. Thirdly,
it assisted in examining the factors that contribute to the intention to drive unsafely at
level crossings as well as self-reported driving behaviour. As such, it assisted in
explaining why personal, social and environmental factors are important in the future of
both engineering and human factor solutions to improving level crossing safety.
6.2 STUDY AIMS AND RESEARCH QUESTIONS
With the assistance of Fishbein’s Integrated Model of Behaviour Change (IM), an
exploratory investigation of three different road user groups was achieved. Two aims and
193
eight research questions were addressed in the final study of the research program. This
summary outlines the findings relevant to each of the aims and research questions.
The first aim was to examine what changes in driving intention and reported
driving behaviour are influenced by exposure to the intervention message.
Research Question 1: Does exposure to the intervention message produce safer
driving intention and self-reported driving behaviour at level crossings?
The second aim was to examine the personal, social and environmental factors that
influence driving intention to perform behaviours related to unsafe driving at level
crossings.
Research Question 2: Is there a relationship between gender and unsafe driving
intention among older and younger drivers?
Research Question 3: Is there a relationship between distance traveled per week
and unsafe driving intention among heavy vehicle drivers?
Research Question 4: Is there a relationship between familiarity and unsafe driving
intention?
Research Question 5: Is there a relationship between attitudes and unsafe driving
intention?
Research Question 6: Is there a relationship between self-efficacy and unsafe
driving intention?
Research Question 7: Is there a relationship between subjective norms and unsafe
driving intention?
Research Question 8: Is there a relationship between beliefs about environmental
constraints and driving intention?
194
6.3 INTERVENTION DEVELOPMENT AND IMPLEMENTATION
6.3.1 Overview
It is well-known that during the past few decades, road safety advertising has
typically had an emphasis on highly graphic (shock style) campaigns such as those used
for drink driving. However, with research now indicating that this style of advertising is
largely ineffective (Christie, 2002), advertising that conveys information to viewers in
order to change beliefs is most likely to be more effective. As discussed earlier in Chapter
Two (Literature Review), it can be understood that some education campaigns that have
typically been developed in a haphazard manner, may have limited effectiveness in
improving road safety. However, it has also be identified that theoretically grounded
campaigns developed in accordance with research and targeting specific road safety
issues can provide a more effective means of risk management. Although the current
project was not a ‘campaign’ per se but rather single interventions, the development and
assessment of educational interventions is just as important in building such campaigns.
The IM (Fishbein, 2003) was utilised in the development of these targeted radio road
safety intervention messages that aimed to change beliefs, attitudes (and subjective norm
and perceived control) and intentions, whilst acknowledging the influence of the
environment (at both the level crossing site level as well as the social
network/community/organisational level) and skills and abilities of motorists.
The interventions were designed in the format of a pilot radio road safety
advertisement, as this medium was found to be one of the most acceptable to each of the
road user groups as identified in the formative research undertaken in Study Two. This
medium was also used as it was within the budget of the project and was likely to be
applied in the ‘real-world’. Radio has been found to be an effective medium for instilling
road safety awareness among drivers (Wong et al., 2004). The radio messages were used
as a one-off targeted awareness raising intervention for each road user group. Each radio
message had a 90 second duration. It was intended that these interventions may be used
in the future by government agencies in support of other mass media campaigns (such as
television, billboards etc) and police enforcement, with the view that campaigns be
195
targeted towards specific road user groups (such as heavy vehicles, older and younger
drivers).
The content of the radio message was different with all three road user groups (see
Appendix 7). However, what was consistent between each of the advertisements was the
purposive non-focus on detection or legal sanctions. As the likelihood of detection at any
level crossing is extremely low in most parts of Australia, focusing on this enforcement
countermeasure would most certainly not be a deterrent to any road user group.
Crossings rarely have police presence or are monitored by cameras (such as red-light
cameras at major intersections) due largely to level crossing fatalities being much lower
than other road safety issues such as drink driving, seat-belt wearing or speeding. There
are only a few level crossings scattered throughout capital cities in Australia that have
video and digital still cameras installed, and even fewer that use such devices for
applying fines and penalties to drivers. Without such enforcement activities to support
educational interventions, it is argued that campaigns used by government agencies will
for the most part be fruitless (Donovan, 1999). This is one of the pitfalls in level crossing
safety research.
6.3.2 Heavy vehicle drivers
The content used for the heavy vehicle radio advertisement was based on the
Victorian level crossing campaign ‘Don’t Risk It’ developed by the Department of
Infrastructure. This campaign provided information for heavy vehicle drivers as well as a
variety of other road user groups. However, information provided for heavy vehicle
driving was supported by the formative research conducted in Study Two. This
formative research indicated that heavy vehicle drivers face three key safety issues at
level crossings: (1) stopping distances, (2) short-stacking and (3) visibility. Therefore,
the information provided to heavy vehicle drivers in this ‘Don’t Risk It’ campaign was
used for ‘suggested actions’ in the intervention advertisement for the heavy vehicle
group.
196
6.3.3 Older drivers
Older drivers were provided with a radio advertisement based on the premise that
they have a wealth of driving experience, but that all level crossings are in some way
different from the next and unfamiliar crossings may be confusing for some older drivers.
Additionally, it acknowledged that older motorists may be vulnerable to being pressured
by other road users, and that it is essential to follow the road rules at level crossings even
if being pressured to take risks. This road user group was also informed that although
they are often very conscious of day-to-day risks, sometimes it can be the unexpected
danger, the once in a lifetime event, that they are not prepared for. Informative research
with this group in Study Two indicated that older drivers believe that they always ‘do the
right thing’ and ‘we don’t take risks’. Many of the drivers that participated in the focus
groups in this formative research phase stated that it is ‘other drivers’ that may pressure
them into having to take risks. Hence, to build a rapport with drivers that listened to this
intervention was the initial step before providing them with actions for maintaining their
safe driving behaviour.
Participants randomly allocated to the older driver control group were also provided
with a road safety message (vision and driving). It was decided that the control group
should receive information about a separate road safety issue that was relevant to this age
group. The reason for this was two-fold. Firstly, it became apparent from conversations
whilst recruiting participants for this group that participants were expecting to be
involved in important road safety research, with part of this being listening to pilot radio
advertisements. Secondly, discussions with experts in the field of road safety suggested
that if older drivers feel that they are contributing to the research (such as pilot radio
advertisements) then they will continue to be participants (i.e. limit the risk of attrition).
Therefore, all participants listened to either information about driving at level crossings
(intervention group) or vision and driving (control group).
6.3.4 Younger drivers
Informative research with younger drivers found that this road user group was
particularly concerned about harming either their friends or their motor vehicle if
197
involved in a road crash or level crossing collision. Many participants in the focus
groups conducted during the formative research, stated that they were however willing to
take risks at crossings to save time. The intervention for this group was centered on the
themes of: (1) knowing the different types of crossings, (2) knowing that trains are not
able to stop in time due their speed and mass and (3) waiting at level crossings whilst a
train is going through may take an extra minute but might save their life. Such themes
were found to be prominent with participants in the formative research phase. The
method for delivery of the intervention was an online survey in which participants
listening to the radio ad either about level crossing safety (intervention) or wearing of
seat belts (control). Each participant had to click on the file to open and listen to the radio
advertisement. This was provided immediately to participants after they had completed
the first questionnaire online.
6.4 METHOD
6.4.1 Research design
There were three parts to this final study. Part One involved the development of
road safety radio advertisements specific for each of the three road user groups. Part Two
involved the investigation of the present context of unsafe driving behaviour at level
crossings. A questionnaire (Time 1) was completed by members of each of the three road
safety groups identified as being ‘at risk’ of a vehicle-train collision. This questionnaire
was then subsequently followed by Part Three of the study. Part Three involved trialing a
pilot road safety radio advertisement using an intervention and control methodology. For
the older and younger driver groups, the design was experimental as participants were
randomly allocated to either the intervention or control group. A post-test questionnaire
(Time 2) was then completed for Part Three by members of each of the three road user
groups. Although for the heavy vehicle driver group the intention was to also use a
quasi-experimental intervention and control design in which there is no randomisation to
intervention or control groups, difficulties arose over time with the recruitment of this
road user group and therefore no control group was able to be recruited. The greatest
challenge was the poor involvement of companies in the study. Although the project
198
sought the assistance of the Queensland Trucking Association and actively approached a
multitude of trucking companies, only five agreed to participate. This was an extremely
disappointing response rate, particularly with the months of effort in approaching
companies and offering incentives to participants. The exact reasons for refusal by
companies to participate in this research remain unknown. However, from discussions
with some trucking company managers that refused to participate, there appeared to be a
sense of either level crossing safety not being a priority safety area for their employees or
that disclosure of employee driving practices could leave their company open to scrutiny.
Other limitations of recruiting from the heavy vehicle industry are discussed in greater
detail later in the chapter.
6.4.2 Questionnaire measures
6.4.2.1 Overview
Two questionnaires were developed for each of the three road user groups: Time 1
(pre-intervention) and Time 2 (post-intervention). Both of the questionnaires was slightly
different for the three road user groups and is provided in Appendix 8. Tables 17 and 18
below illustrate the differences between the Time 1 questionnaire give to the different
road user groups. Questionnaires were piloted on members from each of the road user
groups (i.e. older and younger drivers, and truck drivers).
Each section and instrument of the questionnaires is discussed following the two
tables.
199
Table 17: Items and scales included in the T1 questionnaire Older Drivers
Younger Drivers Heavy Vehicle Drivers
Socio-demographic
Gender, age, employment status, postcode, driving experience, licence restrictions
Gender, age, employment status, postcode, driving experience
Gender, age, employment status, shift work status, postcode, driving experience, licence type
Driving Patterns Frequency and distance travelled per week
Frequency and distance travelled per week
Frequency and distance travelled per week
Health and Driving Self-Assessment
Conditions that may affect driving ability
n/a n/a
Driving Assessment n/a High risk driving behaviours
n/a
History of Road Crashes Involvement in road crashes and contributing factors
Involvement in road crashes and contributing factors
Involvement in work-related road crashes and contributing factors
General Driving Behaviour Modified Driver Behaviour Questionnaire (DBQ)
Modified Driver Behaviour Questionnaire (DBQ)
Modified Driver Behaviour Questionnaire (DBQ)
Level Crossing Driving Behaviour (Self) – past 6 months and next 6 months
Frequency of driving at different types of level crossings and frequency of risk taking behaviours and future intention to risk take
Frequency of driving at different types of level crossings and frequency of risk taking behaviours and future intention to risk take
Frequency of driving at different types of level crossings and frequency of risk taking behaviours and future intention to risk take
Level Crossing Driving Behaviour (Important Others)
Driving behaviour of friends and family
Driving behaviour of friends and family
Driving behaviour of friends and colleagues
Influencing Factors of Driving at Level Crossings
Beliefs of factors that influence self-driving at level crossings
Beliefs of factors that influence self-driving at level crossings
Beliefs of factors that influence self-driving at level crossings
Environmental Constraints
Environmental conditions that affect driving at level crossings
Environmental conditions that affect driving at level crossings
Environmental conditions that affect driving at level crossings
Road Rules Knowledge of level crossing rules
Knowledge of level crossing rules
Knowledge of level crossing rules
Level Crossing Collision Perception of likelihood of being involved in a RLX collision
Perception of likelihood of being involved in a RLX collision
Perception of likelihood of being involved in a RLX collision
200
Table 18: Items and scales included in the T2 questionnaire Older Drivers
Younger Drivers Heavy Vehicle Drivers
General Driving Behaviour
Modified Driver Behaviour Questionnaire (DBQ)
Modified Driver Behaviour Questionnaire (DBQ)
Modified Driver Behaviour Questionnaire (DBQ)
Level Crossing Driving Behaviour (Self) – past 6 months and next 6 months
Frequency of driving at different types of level crossings and frequency of risk taking behaviours and future intention to risk take
Frequency of driving at different types of level crossings and frequency of risk taking behaviours and future intention to risk take
Frequency of driving at different types of level crossings and frequency of risk taking behaviours and future intention to risk take
Level Crossing Driving Behaviour (Important Others)
Driving behaviour of friends and family
Driving behaviour of friends and family
Driving behaviour of friends and colleagues
Influencing Factors of Driving at Level Crossings
Beliefs of factors that influence self-driving at level crossings
Beliefs of factors that influence self-driving at level crossings
Beliefs of factors that influence self-driving at level crossings
Environmental Constraints
Environmental conditions that affect driving at level crossings
Environmental conditions that affect driving at level crossings
Environmental conditions that affect driving at level crossings
Road Rules Knowledge of level crossing rules
Knowledge of level crossing rules
Knowledge of level crossing rules
Level Crossing Collision Perception of likelihood of being involved in a RLX collision
Perception of likelihood of being involved in a RLX collision
Perception of likelihood of being involved in a RLX collision
Radio Message (Intervention or Control)
Recall of message, slogans and information, likelihood that message influenced driving behaviour
Recall of message, slogans and information, likelihood that message influenced driving behaviour
Recall of message, slogans and information, likelihood that message influenced driving behaviour
6.4.2.2 Demographics
A variety of questions were included in the Time 1 questionnaire to determine each
participant’s age, gender, employment status, licence type and/or conditions (e.g.
corrective lenses and/or medical condition requiring a medical certificate), and postcode.
The Rural, Remote and Metropolitan Areas (RRMA) classification was used, which is a
geographical classification based on statistical local areas (SLAs) and allocates each SLA
in Australia to a category based on population numbers and an index of remoteness
(Department of Primary Industries and Energy, 1994). The structure of the RRMA
classification is displayed in Table 19 below.
201
Table 19: Structure of the Rural, Remote and Metropolitan Classifications Zone Classification Category Metropolitan 1 Capital cities 2 Other metropolitan centres (urban centre population > 100,000 Rural 3 Large rural centres with populations 25,000 – 99,000 4 Small rural centres with population 10,000 – 24,999 5 Other rural areas with population < 10,000 Remote 6 Remote centres with population > 5,000 7 Other remote areas with population < 5,000
Younger drivers were also asked about their highest educational level attained.
Additional to employment status, heavy vehicle drivers were also asked about shift work,
the length of time having held their licence to drive a heavy vehicle, and the hours and
distance driver on an average week. Therefore, the questionnaires for each road user
group were slightly different with regards to socio-demographic information.
6.4.2.3 Crash history
Each of the three road user groups was asked questions about their crash
involvement during the past three years. Heavy vehicle drivers were specifically asked
about their crash involvement whilst driving a truck. For those participants that reported
having been involved in a road crash, they were asked questions about the number of
crashes, who was at fault, any factors that may have caused the crash and if any person
was injured or died as a result of the crash.
6.4.2.4 Medical conditions
Older drivers were asked a range of questions related to what medical conditions
they have been diagnosed with. This instrument was developed by reviewing the
literature on older drivers and making some assumptions about influences that medical
conditions may have on driving ability. As this instrument is made up of medical
conditions, reporting the Cronbach’s alpha coefficient for the 17 items is unnecessary.
6.4.2.5 Knowledge of level crossing road rules
Four questions were asked about knowledge of level crossing road rules at each
time point. It was necessary to gain an understanding about the road rules that drivers
were unaware of. The questions included: what twin flashing lights mean, whether boom
202
gates stay down a lot longer than normal intersections, what yellow hatching road
markings mean, and if there are fines for failure to stop at a level crossing.
6.4.2.6 General driving behaviour
The measure of driving behaviour used for each road user group was a modified
version of the Driver Behaviour Questionnaire (DBQ). This questionnaire has become
widely used in examining self-reported driving behaviour (Lajunen, 2003). This
instrument has been extensively used in a variety of road safety research in many
countries throughout the world (Davey et al., 2007). Reason et al’s (1990b) original DBQ
focused on two distinct behaviours: errors and violations. Errors are defined as driving
actions that are not planned, whilst violations are defined as deliberate deviations from
safe driving practices (Reason, 1990b). However, both behaviours are potentially
dangerous in the context of driving (Lajunen, 2004b). ‘Slips and lapses’ was later
identified as focusing on attention and memory failures, however these were not
considered to affect driving safely significantly. Lajunen and Summala (2003) suggest
that errors are more serious mistakes such as failure of observation or misjudgments,
while lapses are more likely to be associated with memory and attention difficulties.
Lawnton et al (1997a) revised the original DBQ to differentiate violations as either
aggressive or ordinary. Aggressive violations are considered to be associated with an
interpersonally aggressive element, whilst ordinary violations do not aspire to being
aggressive, but are still considered deliberate violations (Lawton, 1997a). Research has
confirmed different loading of factors. Aberg and Rimmo’s (1998) research has
confirmed the three factors (errors, lapses and violations), while Sullman et al’s (2002)
research on New Zealand truck drivers confirmed four factors (errors, lapses, ordinary
violations and aggressive violations).
A modified version of the DBQ was used in both questionnaires (i.e. Time 1 and
Time 2) for each of the three road user groups. As there is no known research that has
utilised the DBQ for level crossing driving, it was considered necessary to include
additional items. For older drivers and younger drivers, the instrument consisted of 24
items. This instrument was developed by researchers at the Centre for Accident Research
and Road Safety – Qld, based at the Queensland University of Technology. The internal
203
consistency of these modified DBQ’s for younger and older drivers was examined
through calculating Cronbach’s alpha reliability coefficient. Similar to previous research
with younger drivers (Ulleberg & Rundmo, 2003) and older drivers (Parker, 2000) the
instrument appears to exhibit relative internal consistency (0.802 and 0.801 respectively).
As professional truck drivers are often under time pressure, it was necessary to
include a modified version of the DBQ to assess its usefulness in determining
relationships such as crash involvement (Davey et al., 2007). Research by Sullman et al
(Sullman, 2002) was reviewed and although the 28-item DBQ used was considered
adequate for other research purposes, it was deemed to not be extensive enough for the
current study. A total of 41 items was included in the modified DBQ for heavy vehicle
drivers (24 items included in older and younger driver questionnaires as well as 17 items
regarding professional driving). This instrument was also developed by researchers at the
Centre for Accident Research and Road Safety – Queensland, based at the Queensland
University of Technology. The internal consistency of this modified DBQ for heavy
vehicle drivers was examined through calculating Cronbach’s alpha reliability
coefficient. Similar to previous research with heavy vehicle drivers (Sullman, 2002), the
instrument appears to exhibit relative internal consistency (0.834).
Factor analysis was also administered on both instruments (41-item and 24-item).
Principal components analysis with varimax rotation was implemented to determine the
factor structure both of these modified DBQ’s. One item ‘have one or two alcoholic
drinks before driving’ had to be removed from the factor analysis for the heavy vehicle
driver instrument (41-item) as there was no variance found when computed by SPSS.
With this item removed, a twelve-factor solution accounted for 23.38% of the total
variance. Different methods of factor extraction and rotation were undertaken, however
none yielded any better factor structure results. According to Ozkan, Lajunen and
Summala (2006):
It is likely that the number of items, various versions of DBQ, different sampling
strategies (i.e., postal survey), different target populations (i.e., elderly drivers),
driving context (i.e., work and leisure), different traffic cultures and environment
all influence the stability of the DBQ structure.
(p387)
204
6.4.2.7 Self-reported driving behaviour at level crossings
Participants were asked a range of questions related to driving behaviour at level
crossings that were considered by the research team to be high risk driving. These
statements were developed from a combination of formative research, review of the
literature as well as observations of driving behaviour at level crossings. A 20-item
instrument was developed for both younger and older drivers for both questionnaires (i.e.
Time 1 and 2). Unfortunately, the heavy vehicle driver instrument had only 19 items as
one item (seen police enforcing the road rules at rail crossing) was unintentionally deleted
during questionnaire compilation. Participants were asked to score how they had driven
during the past six months at level crossings (for Time 1) and past month (Time 2). The
scale ranged from 0=Not at all likely to 5=Very likely. The alpha coefficient for both the
younger driver (0.978) and heavy vehicle driver (0.758) instruments were considered
acceptable, while the older driver (0.655) instrument was slightly less than the acceptable
value of 0.7.
6.4.2.8 Intended driving behaviour at level crossings
As well as reported driving behaviour, behaviour intention at level crossings was
also measured. The same scale for reported behaviour was used for intention. However,
participants were asked how likely they would behaviour in the next six months of
driving (0=Not at all likely to 5=Very likely). The alpha coefficient for older drivers
(0.626) was a little low, while the younger drivers (0.925) and heavy vehicle driver
(0.858) instruments were considered acceptable. Like the scale for self-reported driving
behaviour at level crossings, unfortunately the heavy vehicle driver instrument had only
19 items instead of 20. One item ‘seen police enforcing the road rules at rail crossing’
was unintentionally deleted during questionnaire compilation.
6.4.2.9 Attitudes towards driving at level crossings
A direct measure of attitude was derived from 8 items from the 15-item instrument
bipolar scales (-3 to +3). The first 8 items in this instrument directly measured attitude
towards driving at level crossings, while the last 7 items directly measured control. Each
of the bipolar scales was recoded in SPSS using unipolar (1-7) scales. Participants were
205
asked to complete statements by rating pairs of adjectives whilst driving at level crossings
(e.g. bad-good, more confusing-less confusing). Cronbach’s alpha statistics for the 8
items in this 15-item instrument were above the acceptable level of 0.7 for both heavy
vehicle drivers (0.865) and older drivers (0.892), however was slightly lower than the
acceptable level of 0.7 for younger drivers (0.622).
6.4.2.10 Self-efficacy/perceived behavioural control of driving at level crossings
A direct measure of self-efficacy/perceived behaviour control was derived from the
last 7 items in the 15-item instrument that used bipolar scales (-3 to +3). The first 8 items
in this instrument directly measured attitude towards driving at level crossings, while the
last 7 items measure control. Each of the bipolar scales was recoded in SPSS using
unipolar (1-7) scales. Participants were asked to complete statements by rating pairs of
control adjectives whilst driving at level crossings (e.g. up to me-not up to me, dependent
on other motorists-dependent on me). Cronbach’s alpha statistics for the 7 items in this
15-item instrument were above the acceptable level of 0.7 for heavy vehicle drivers
(0.763), older drivers (0.705) and younger drivers (0.733).
6.4.2.11 Subjective norms other important others
A 14-item instrument for directly measuring the construct of subjective norm was
developed for each of the road user groups. The Cronbach’s alpha statistic for internal
consistency for younger drivers (0.704) was above the acceptable 0.7 level, however both
the heavy vehicle (0.610) and older driver (0.579) instruments showed lower internal
consistency.
6.4.2.12 Perceived risk of a level crossing collision
A scale was developed for participant’s perception of the likelihood that they would
be involved in a collision with a train at any point in the future. The scale was from 0
(not at all likely) to 5 (very likely).
206
6.4.2.13 Environmental constraints whilst driving at level crossings
Participants were asked a range of questions about their beliefs of specific
environmental and engineering factors at level crossings. The frequency each of the
driving situations occurred was recorded from 1=Never to 5=Always. Heavy vehicle
drivers were asked a total of 13 questions specifically for driving a truck, while younger
and older drivers were asked only nine (9) questions. Each of the instruments had an
acceptable alpha coefficient with the younger driver instrument being the highest (0.818),
followed by they heavy vehicle instrument (0.806) and older driver instrument (0.741).
6.4.2.14 Familiarity with different protection systems
Participants were also asked how frequently they drove through different level
crossing protection systems (i.e. active crossings with boom gates, passive crossings with
flashing lights and passive crossings with stop or give-way signs only). Response items
included: never, once a year, twice a year, monthly, weekly or daily. Familiarity with
each of the types of protection systems was then recoded to participants being either
familiar or unfamiliar drivers.
6.4.3 Pilot testing
Time 1 (Part One) questionnaires for the three road user groups were piloted after
compiling the findings from the formative research and available literature in the area.
Road and rail safety experts in the field as well as a selection of participants from each of
the road user groups assisted with the pilot testing. Managers from the participating truck
depots piloted the questionnaires, while drivers aged 60 plus years assisted in piloting the
older driver questionnaire. Young drivers (aged 17–24 years) that worked with the
researcher served as participants for piloting the young driver questionnaire. Items were
refined based on initial responses and interpretations. As the Time 2 questionnaire
contained almost identical items to that of Time 1, it was deemed unnecessary to pilot
this questionnaire as well.
207
6.4.4 Ethical considerations
Ethical clearance for data collection for this final study was gained from the
Queensland University of Technology Human Research Ethics Committee (QUT Ref.
No. 3550H). All participants were informed that:
• Their participation was strictly voluntary and their confidentiality was assured;
• Any data collected will be unidentifiable and purely for research purposes;
• Personal data would only be accessibly by members of the research team;
• There would be no direct benefits to them or any risks by participating in the
study;
• If they would like further information they could contact the chief investigator;
and
• If they had any concerns or complaints about the ethical conduct of the project
they should contact the research ethics officer (telephone number provided).
6.4.5 Sample size calculations
Typically a number of factors must be known or estimated to calculate sample size.
These include (1) the effect size (usually the difference between two groups); (2) the
population standard deviation (for continuous data); (3) the desired power of the
experiment to detect the postulated effect; and (4) the significance level (Sudman, 1976).
However, in this study, there is no prior research that can inform the estimate of the
population standard deviation of any of the variables used. Therefore, a statistician was
involved in the calculation of the sample size for this study. Sudman (1976) suggests that
a minimum of 100 participants is needed for each major group in a sample and for each
minor sub-group, a sample of 50 participants is necessary. To allow detection of a
difference in driving behaviour at level crossings between intervention and control
groups, as well as cost considerations, a sample size of 200 for each of the road user
groups was identified as being an appropriate sample. It was hoped for both the older and
younger drivers that this sample would be split in halves for regional and metropolitan
areas, and then split again in halves for intervention and control participants (e.g. 50
intervention regional older drivers, 50 control regional older drivers etc). For the heavy
208
vehicle drivers, it was known that this group would be the most difficult road user group
to recruit from (advice from other researchers) and the aim was to recruit 200 drivers in
total, while ensuring that all companies recruited traveled in both regional and
metropolitan areas. Limitations of recruiting from the heavy vehicle industry are
discussed in detail in ‘study limitations’ (6.7).
6.4.6 Recruitment strategy
6.4.6.1 Older drivers
Drivers aged 60 years and over were recruited through the Royal Automobile Club
of Queensland (RACQ). RACQ was approached to assist with recruitment of 200 drivers
in both regional and urban areas. A letter from the General Manager for External
Relations was sent to a randomly selected list of 1000 older drivers (500 urban and 500
regional) during June 2006. This letter supported the research being conducted but did
not state it was specifically on the subject of railway level crossings. Although bias
exists in all research, it was important that the research limited the amount of possible
bias and therefore the omission of the exact nature of the research was deemed by
statisticians as being vital. The letter stated that if RACQ members would like to
participate, they should telephone or email the project officer to leave their contact
details. The project officer was employed by the larger project as the contact person for
all external relations as well as assisting with data collection.
6.4.6.2 Young drivers
Young drivers aged 17-24 years were also recruited through the Royal Automobile
Club of Queensland (RACQ). As the RACQ assisted with the recruitment of older
drivers, it was deemed an efficient and inclusive means to also recruiting young drivers.
Like the older driver recruitment strategy, a letter from the General Manager for External
Relations was also sent to a randomly selected list of 1000 young drivers (500 urban and
500 regional) during December 2006. This letter supported the research being conducted
but did not state it was specifically about driving at level crossings. To ensure there was
209
limited bias, the omission of the exact nature of the research was deemed by statisticians
as a vital part of the recruitment strategy.
6.4.6.3 Heavy vehicle drivers
The Queensland Trucking Association (QTA) assisted with recruitment for this
research. A list of ten major truck companies that were likely to participate was provided
by the QTA. Each of the trucking companies were contacted and informed about the
project and invited to participate. As an incentive for participation, safety managers were
told that their company would be provided with a ‘risk profile’ of the aggregate data from
their drivers. This incentive was considered to be a motivating factor for the participating
companies. Additionally, each of the heavy vehicle drivers was informed that if they
completed both questionnaires then they would be mailed two movie tickets.
6.4.7 Sampling method
6.4.7.1 Older and younger drivers
A non-probability method (voluntary response sampling) was used for both the
older and younger driver samples. As members of both these road user groups were
invited to participate through a mail out letter from the RACQ, they voluntarily decided
to participate in the research. The advantage of this type of sampling is that it is efficient
for collecting large amounts of information and flexible in collecting information
regarding attitudes, beliefs, and reported and intended behaviours. The disadvantage of
this method is that the sample may include individuals with strong opinions about the
issue, and can therefore be biased (voluntary response bias). However, given that
participants were informed that they would receive two movie tickets for completion of
both questionnaires, this may have been very appealing to invited participants and
therefore the sample may not have just included individuals with strong opinions about
level crossings. With a limited budget for recruitment, this method of sampling was found
to be the most efficient way of recruiting participants quickly.
210
6.4.7.2 Heavy vehicle drivers
A non-probability sampling method (purposive/judgmental) was used for the
collection of data from this industry. Although probability sampling is the preferred
method by statisticians and social scientists alike, in the ‘real-world’ of research it is not
always possible to use a probability sample. The advantage of this type of sampling is
that it is less expensive, efficient and easy to implement (Foreman, 1991). Purposive
sampling can also be very useful when a target sample needs to be reached quickly and
when sampling for proportionality is not the primary concern (Colman, 1995). For the
purposes of the current mode of inquiry, this type of sampling was within the budget of
the project as well as the most appropriate and suitable for sampling within the trucking
industry.
However, the limitations of this sampling method must be addressed. One major
limitation of this type of sampling is the introduction of bias due to the sample pattern
(Foreman, 1991). Therefore, non-probability samples cannot depend upon the rationale
of the theory of probability; however they can serve a purpose when there is reliable
information about the sample (or site). Assistance from the Queensland Trucking
Association (QTA) offered extra validity of the data as there was professional judgment
of the companies involved. The other limitation of this type of sampling is that it limits
the usefulness of the data for statistical interpretation (Tilley, 1990, Colman, 1995).
However, it does not necessarily mean that non-probability samples are not representative
of the population involved. It was the not the intention of this research to work out what
proportion of the trucking population gives a particular response, but rather to obtain an
idea of the range of responses on the attitudes, beliefs and intentions of driving at level
crossings. With the involvement of small, medium and large trucking companies in the
research, this assisted with ensuring the sample was representative.
6.4.8 Procedure and response rate
6.4.8.1 Older drivers
Of the 1000 letters that were sent by RACQ, 186 members telephoned the project
officer to provide contact details for sending the first questionnaire (Time 1). This
211
response rate equated to 18.6%. Participants were then randomly selected, using a table
of random numbers, to either the intervention or control group. The first questionnaire
was sent in mid-July 2006 to the 186 older drivers that agreed to participate. This
questionnaire also included an Information Sheet. Of the 186 questionnaires sent at Time
1, 152 completed questionnaires were returned. The response rate for Time 1 was
therefore 15.2% (response rate of 81.7% for those that originally agreed to participate).
Six (6) weeks after receiving the completed Time 1 questionnaire from participants, a
research assistant with a degree in Psychology telephoned participants and played the
road safety radio advertisement intervention or control message. Telephoning took place
both during the day and in the evening to capture as many participants as possible.
During this telephoning period (one week), 137 participants were able to be contacted.
The response rate was therefore 13.7% (response rate of 73.7% for those that originally
agreed to participate). The intervention and control messages are provided in Appendix 7.
The Time 2 questionnaire (post-intervention) was sent to the 137 participants that
listened to either the control or intervention radio advertisement. This was sent 4 weeks
after being exposed to the advertisement. There were 109 participants that completed and
returned this second questionnaire, which was a response rate of 10.9% (response rate of
58.6% for those that originally agreed to participate).
6.4.8.2 Younger drivers
In December 2006, a letter was sent from the RACQ supporting this research was
sent by post to 1000 of its younger members aged 17-24 years. Members were informed
of an online survey (two in total) that would take approximately 20 minutes to complete.
This survey was developed by an external company that had been employed by the
Centre for Accident Research and Road Safety – Qld (CARRS-Q) previously. Incentives
for completing both online surveys were two movie tickets. The sample of participants
was divided into intervention and control groups by month of birth. With online surveys,
randomisation of participants is often difficult and therefore this method was deemed
most suitable. Therefore, participants born between January and June were allocated into
the intervention group. Participants born between July and December were allocated into
the control group.
212
In December 2006, 149 younger drivers accessed the online survey and completed
the first questionnaire. There were 100 participants born between January and June
(intervention group) and 49 born between July and December (control group). This
equated to a response rate of 14.9%. At the completion of this first online questionnaire,
each participant listened to an online audio file either about safe driving at level crossings
or the wearing of seatbelts. Participants of the Time 1 questionnaire were sent an email 4
weeks after completion of the first questionnaire informing them it was time to complete
the second questionnaire (Time 2). The total response rate for this second questionnaire
was 88 (30 control group and 58 intervention group). A reminder email was sent two
weeks after this email to remind Time 1 participants to complete the second questionnaire
in order to receive the movie tickets. The response rate for this questionnaire from the
Time 1 questionnaire was 59.1%.
6.4.8.3 Heavy vehicle drivers
Five (5) companies that were invited to participate through the Queensland
Trucking Association agreed to participate in the research. The other five (5) companies
stated industrial relation issues and poor staffing as limiting them from participating.
During July through to September 2006, data collection sessions were arranged with each
of the five (5) participating truck companies. As well as the author, a research assistant
with a degree in Psychology collected the data. An information sheet was provided to
each participant with the Time 1 (pre-intervention) questionnaire. Consent was implied
with completion of the questionnaire. Time 1 questionnaires were completed by
participants in a designated quiet area (such as the lunch room) at the truck depot. Each
participant provided mailing details to send the post-intervention (Time 2) questionnaire
to. This method was deemed the most suitable for the trucking companies as it ensured
confidentiality of questionnaires. Forty-five (45) heavy vehicle drivers turned up to the
lunch room and said they would be happy to participate. Each driver was given an
Information Sheet to read and consent was implied by completing the questionnaire. All
participants were informed that no information could be traced by their company and that
the data would become aggregate data to be analysed.
213
Once participants had completed the Time 1 questionnaire, they listened to the
intervention message. The intervention was a pilot radio road safety message specific for
driving at railway level crossings. As participants completed the questionnaire at different
times, often due to varying literacy levels, it was important to ensure that other
participants could not overhear the intervention being played. Some truck drivers
completed the questionnaire in 15 minutes while others took up to 45 minutes.
Therefore, an iPoD with earphones was used for each participant to listen to the radio
message. Each participant was then informed they would be receiving another
questionnaire in the mail. Depending on arrangements made with each of the truck
company’s operations manager, some participants were informed that they would receive
two movie tickets in the mail once both questionnaires were completed.
The Time 2 questionnaire (post-intervention) was sent to the 45 truck drivers that
completed the Time 1 questionnaire. This questionnaire was sent 4 weeks after
participants were exposed to the radio advertisement. A follow-up reminder was sent 2
weeks later to those drivers that did not return the Time 2 questionnaire. A total of 11
drivers completed and returned the Time 2 questionnaire. This response rate was
therefore 24.4%.
Table 20: Response rates compared across road user groups Road User Group
Letter sent
T1 sent
T1 received
T2 Sent
T2 received
Older drivers 1000 186 152 137 109 Younger drivers 1000 - 149 - 88 Heavy vehicle drivers - - 45 45 11
6.4.9 Data management
Data was entered using the Statistical Package for the Social Sciences (SPSS)
version 15.0. Before commencement of data analysis, all databases were cleaned to check
for accuracy of data entry, missing values and outliers. For both the heavy vehicle
drivers and older drivers, there was minimal missing data. However, for the younger
driver group, one instrument, ‘subjective norms’ had a significant number of missing data
for the first three items and therefore these three items were excluded from both Time 1
and Time 2 to ensure equal comparisons. As this road user group completed their
214
questionnaires online, it is possible there was electronic saving problems for the page in
which this instrument was displayed.
The level of significance for measurement was set at p<.05. A more rigorous
significance level (α = .01) was used for any post-hoc comparisons, to guard against
inflation of the Type 1 error rate (i.e. rejecting the null hypothesis when it is true).
6.4.10 Dependent variables
One variable was selected to act as the dependent variable in order to operationalise
Fishbein’s integrated model: driving intention at level crossings (Time 1). Although
intentions are not a precise predictor of future behaviour, they are the primary
determinant of behaviour and signify a preparedness to engage in particular behaviours
(Fishbein et al., 2001). Since both the older and younger driver samples utilised an
experimental design (pre and post-testing), the variable ‘intention’ (Time 1) was used as
it was not contaminated by the effects of exposure to the intervention.
6.4.11 Data analysis
A combination of descriptive statistics, parametric testing and repeated measures
was undertaken for this final study. For the heavy vehicle group, both Analysis of
Variance (ANOVA) and paired sample t-tests were performed. Since the entire sample
of heavy vehicle drivers received the intervention (no control group), then paired sample
t-tests were the most appropriate statistic. Given that both the younger and older driver
groups were randomly allocated to either the intervention or control group, split-plot in
time analysis were used instead of paired sample t-tests to test the effect of exposure to
the intervention. Additionally, hierarchical regression was performed on these two
groups to predict which variables influence intention to drive safely at level crossings.
These regressions used Time 1 variables only, as the influence of exposure to the
intervention would have contaminated the prediction of intention.
One major assumption of the ANOVA is that each group is an independent random
sample from a normal population. For the heavy vehicle sample, it was impossible to
randomly select groups (i.e. companies) from the normal population as companies needed
to give their permission for their drivers to participate in the research. Many trucking
215
companies that were approached declined to participate in the research and therefore non-
probability purposive sampling was used. Due to the variety of trucking companies that
participated in the research, the sample was deemed representative (i.e. characteristics
correspond to, or reflect, those of the original population or reference population). An
alpha level of .05 was used for all ANOVA’s and r was calculated as the effect size
(Rosenthal, 1991).
A paired sample t-test is used to verify the significance of the difference between
two sets of paired data (with no control group). This t-test is used to compare two
population means where there are two samples in which observations in one sample can
be paired with observations in the other sample (Shier, 2004). A typical example of this is
the pre and post-test observations on the same subjects (Colman, 1995). Although
attrition was high for the heavy vehicle sample between Time 1 (n=45) and Time 2
(n=11), statistical advice indicated that paired sample t-tests could be conducted on these
two time points. Additionally, as regression can only be applied to sample sizes larger
than 50, it was not able to be applied to the heavy vehicle dataset. However, hierarchical
regression was able to be applied to both the younger and older driver samples. In
hierarchical regression, independent variables are not inserted into the model all at once,
but rather they are entered into the model in an order specified by the researcher (or
informed by theory). By doing this, each predictor variable (independent variable) can be
assessed in terms of what it adds to the model at its own point of entry (Tilley, 1990).
This type of regression assists in determining which factors/constructs may satisfy or
increase the intention/behaviour relationship.
6.5 RESULTS
6.5.1 Heavy vehicle drivers
6.5.1.1 Company characteristics
Four trucking companies agreed to participate in the current study. Due to
confidentiality agreements with each of the trucking companies, pseudonyms are
216
provided for each company. All of the companies had large depots located in Brisbane.
A profile of each trucking company is provided below.
• ‘Abbotts’ is one of the pioneer companies of integrated logistics in Australia as
well as being an industry leader in the provision of warehousing, transportation
and supply chain management. It has a fleet of over 4200 vehicles traveling to and
from more than 250 sites throughout Australia. This fleet is considered to be a
large size fleet.
• ‘Carters’ provides refrigerated transport services throughout Australia including
local distribution in Sydney and Brisbane. Its fleet consists of over 140 company
owned prime movers, 180 refrigerated vans and 35 rigid trucks. This fleet is
considered to be a medium size fleet.
• ‘Rogers’ is a leading provider of transport and third party warehousing and
distribution services throughout Australia. Roger’s line haul fleet of interstate
prime movers covers all capital cities as well as regional areas in Queensland.
This company fleet consists of around 50 prime movers, 60 flat-top trailers, 15
skel trailers, 18 drop-deck trailers, 15 curtain-sided trailers , 6 prairie wagon
trailers and 100 containers. This fleet is considered to be a medium size fleet.
• ‘Thompsons’ is a smaller private enterprise that provides freight transport
throughout Brisbane, Gympie, Hervey Bay, Maryborough and Bundaberg, as well
as to Sydney. More specific information about this company was not released by
company managers as it was deemed confidential. This fleet is considered to be a
small size fleet.
6.5.1.2 Sample attrition
The attrition rates for participants from each of the companies are illustrated in the
table below. Abbott’s had the best retention rate of participants, while Thompson’s had
the poorest retention rates.
217
Table 21: Company attrition Time 1 Time 2
Company n % n % Abbotts 13 28.9 5 45.5 Thompsons 6 13.3 1 9.1 Rogers 13 28.9 4 36.3 Carters 13 28.9 1 9.1 Total 45 100.0 11 100.0
6.5.1.3 Demographics
At Time 1, the mean age for drivers was 47.4 years (S.D. 9.203), with the youngest
participant being 30 years and the oldest participant being 64 years. By Time 2, the mean
age of participants increased to 51.4 years (S.D. 7.061). The table below compares the
means of the four companies at both time points.
Table 22: Mean age between companies Time 1 Time 2
Company Mean Age Mean Age ‘Abbott’s 47 50 ‘Thompsons’ 50 49 ‘Rogers’ 49 53 ‘Carters’ 45 55
The majority of participants were full-time workers at Time 1 (93.3%), with 42.2%
stating that they work shift work. Most of the participants had a MC (B-double or road
train) licence (66.7%), drove between 4-6 days per week (71.1%), and drove between 51-
72 hours per week (62.2%). Nearly 16% (n=7) of drivers reported driving more than 5000
kilometres per week, while 26.7% reported driving less than 1000 kilometres per week.
A mean of 25.2 years (S.D. 10.724) was observed for number of years licensed, with a
range of 2 to 40 years.
ANOVA’s were also calculated to examine differences between shift workers and
each of the driving instruments. Results are presented in the table below.
218
Table 23: Mean overall scores between shift work and driving at level crossings Instrument Shift Work n df Mean F Sig DBQ Shift work 18 1, 41 0.70 3.398 0.72 No shift work 25 0.55 Self-reported driving behaviour Shift work 19 1, 43 1.31 4.330 0.043 No shift work 26 1.16 Intended driving behaviour Shift work 18 1, 42 1.40 11.984 0.001 No shift work 26 1.08 Subjective norms Shift work 18 1, 42 5.82 0.027 0.871 No shift work 26 5.87 Attitude Shift work 18 1, 42 5.43 0.027 0.871 No shift work 26 5.48 Perceived behavioural control Shift work 19 1, 43 5.81 .495 .486 No shift work 26 5.62 Beliefs about environmental constraints Shift work 19 1, 42 2.36 6.641 0.014 No shift work 25 1.97
A significant effect was found for three of the driving instruments. An effect was
revealed for reported driving behaviour F (1, 43) = 4.330, p <0.05, intended driving
behaviour F (1, 42) = 11.984, p <0.01, and beliefs about environmental constraints F (1,
42) = 6.641, p <0.05. It appears that truck drivers that reported working shifts are
significantly more likely to take risks when driving generally and at level crossings, and
hold beliefs that environmental constraints affect their driving at level crossings.
6.5.1.4 Crash history
Nine participants (20.0%) reported at Time 1 of having been involved in a road
crash whilst driving a truck during the past 3 years. Table 24 illustrates the number of
drivers from each company involved in a road crash during the past 3 years. As can be
seen, there was very little difference between the numbers of crashes from each company,
however, three out of the six Thompson’s participants (50%) reported being involved in a
crash.
219
Table 24: Crash involvement between companies Company Involved in Crash *
(n) Total Participants
(n) Abbott’s 2# 13 Thompson’s 3 6 Roger’s 2 13 Carter’s 2# 13 Total 9 45
N.B. * denotes being involved in a crash during past 3 years. # denotes one participant being involved in 2 crashes during past 3 years.
Of those participants who reported being involved in a crash during the past 3
years, seven (77.7%) reported being involved in one road crash, whilst two participants
(22.3%) reported that they were involved in two road crashes. Interestingly, no
participant who had been involved in a road crash stated that it was their fault. Six
(66.6%) participants stated that the crash was ‘not at all’ their fault, while the other three
participants did not answer the question. Additionally, no participant reported that they
had been involved in a level crossing collision. Table 25 illustrates the seriousness and
type of road crashes that participants reported being involved in.
Table 25: Road crashes during past 3 years Crash seriousness and type n % Seriousness Damage only 7 77.7 Slight injury (to any person) 0 0 Serious injury (to any person) 1 11.1 Fatality (to any person) 1 11.1 Level crossing collision 0 0 Type Head on collision with another vehicle 0 0 Rear-end collision with another vehicle 1 11.1 Angular (i.e. side-on) collision with another vehicle 3 33.3 Collision with a pedestrian 0 0 Collision with another object (e.g. parked car, animal, tree) 1 11.1
Overturned vehicle 1 11.1
Participants were also asked at Time 1 what they believed were contributing factors
to their road crash during the past 3 years, with the scale of 0=Not at all true,
1=Somewhat true and 2=Very true. Four (44.4%) participants stated that they ‘couldn’t
220
see the other vehicle or object’, while two (22.2%) participants stated that they were
‘listening to music or the CB radio’ and two (22.2%) stated that they were ‘distracted by
something outside the vehicle’. Participants could give more than one answer to being a
contributing factor.
Table 26: Contributing factors to road crash
Type n % Been drinking alcohol before driving 0 0 Felt tired 0 0 Driving too fast for the conditions 0 0 Talking to a passenger in the vehicle 0 0 Listening to music or the CB radio 2 22.2 Talking on a mobile phone 0 0 Trying to pick up something from the seat or floor 0 0 Adjusting the radio/cassette/CD or fan/air conditioning 0 0 Checking the instruments (e.g. fuel gauge or speedometer) 1 11.1 Daydreaming 0 0 Talking on the CB radio 1 11.1 Distracted by something inside the vehicle 0 0 Distracted by something outside the vehicle 2 22.2 Couldn’t see the other vehicle / object 4 44.4
N.B. n=participants that reported contributing factor being ‘somewhat true’ or ‘very true’.
6.5.1.5 Knowledge of level crossing road rules
Less than a third of participants (n=14) answered all four questions correctly at the
first time point. Eleven (24.4%) participants answered only two questions correctly at
Time 1, which decreased to 9.1% (n=1) at Time 2. Table 27 below shows the differences
between time points.
Table 27: Knowledge of level crossing rules and facts Time 1 Time 2 Type
n % n % All questions correct 14 31.1 3 27.3 Three questions correct 20 44.4 7 63.6 Two questions correct 11 24.4 1 9.1 No correct questions 0 0 0 0
221
6.5.1.6 General driving behaviour
Mean scores for the modified Driver Behaviour Questionnaire (DBQ) are presented
in the table below. All self-reported mean scores were relatively low, indicating that
participants believe that they are driving safely most of the time.
Table 28: Modified DBQ Time 1 Time 2
Company n Mean n Mean Abbotts 13 .57 5 .77 Thompsons 6 .65 1 .83 Rogers 11 .57 3 .83 Carters 13 .68 1 .32
1=Not at all to 5=Very often
Results of paired sample t-tests indicate that there was no significance difference in
the pre and post-intervention paired data sets for general driving behaviour (i.e. DBQ)
before the intervention (M=0.70, S.D.=0.370) to after the intervention (M=0.73,
S.D.=0.420), t(9) = -.563, p = 0.589.
6.5.1.7 Self-reported driving behaviour at level crossings
Mean scores of the reported driving behaviour at level crossings at Time 1 and
Time 2 is presented in Table 29 below. Carters had the highest mean scores at both Time
1 and Time 2. Results of paired sample t-tests indicate that there was no significance
difference in the pre and post-intervention paired data sets for self-reported driving
behaviour at level crossings of truck drivers before the intervention (M=1.22,
S.D.=0.370) to after the intervention (M=1.20, S.D.=0.242), t(11) = 2.12, p = 0.836.
Table 29: Self-reported driving behaviour at crossings Time 1 Time 2
Company n Mean n Mean Abbotts 13 1.15 5 1.13 Thompsons 6 1.18 1 1.00 Rogers 13 1.15 4 1.24 Carters 13 1.38 1 1.63
1=Not at all to 5=Very often
222
6.5.1.8 Intended driving behaviour at level crossings
Mean scores the intended driving behaviour at level crossings at both Time 1 and
Time 2 is presented in the Table 30 below. Like self-reported driving behaviour, means
were relatively low, with Carters being the highest at both time points. Results of paired
sample t-tests indicate that there was no significance difference in the pre and post-
intervention paired data sets for self-reported intended driving behaviour at level
crossings of truck drivers before the intervention (M=1.26, S.D.=0.497) to after the
intervention (M=1.29, S.D.=0.437), t(11) = -0.112, p = 0.913. Table 30: Intended driving behaviour at crossings
Time 1 Time 2 Company n Mean n Mean Abbotts 13 1.06 5 1.11 Thompsons 5 1.05 1 1.00 Rogers 13 1.24 4 1.50 Carters 13 1.39 1 1.63
1=Not at all likely to 5=Very likely
Bivariate correlations were also undertaken to examine the relationships between
driving intention at level crossings (Time 1) and each of the independent variables
(outcome variables). As shown in Table 31, there was a weak positive correlation
between the driver behaviour questionnaire and driving intention at level crossings (Time
1) [r = .315, p < .05]. This indicated that participants that drive unsafely on the roads
generally are more likely to drive unsafely at level crossings. A significant negative
correlation was found between shift-work and driving intention at level crossings (Time
1) [r = -.471, p < .01]. In other words, participants that worked shift-work were more
likely to hold strong intentions to drive unsafely at level crossings. Additionally, a strong
positive correlation was found for kilometres per week drive and driving intention at level
crossings (Time 1) [r = .405, p < .01] which indicates that truck drivers that drive greater
distances per week have stronger intentions to drive unsafely at level crossings.
Moreover, a significant correlation was found for perceived risk of a level crossing
collision and driving intention at level crossings (Time 1) [r = .630, p < .01]. In other
words, participants that had stronger intentions to drive unsafely at level crossings were
223
more likely to have higher perceptions of risk with being involved in a collision at a level
crossing. As predicted, self-reported driving behaviour (Time 1) was significantly
correlated with intended driving behaviour (Time 1) [r =.647, p < .01]. Since the
sample size at Time 1 (n=45) dropped considerably at Time 2 (n=11), correlating
independent variables with self-reported driving behaviour at level crossings (Time 2)
would have most likely produced unreliable correlation data and therefore excluded from
bivariate correlations.
Table 31: Bivariate correlations between dependent and independent variables
Dependent Variable
Independent variables
Driving Intention at Level Crossings (T1)
Age -.134 Years held licence -.112 Shift-work -.471 ** Kilometres per week drive .405 ** Crashes during past 3 years .017 Driver behaviour questionnaire .315 * Attitudes towards driving at level crossings -.201 Self-efficacy of driving at level crossings -.228 Normative beliefs of others driving at level crossings -.033 Beliefs of environmental constraints whilst driving at level crossings .153 Perceived risk of crash at a level crossing .630 ** Driving intention at level crossings (T1) - Self-reported driving behaviour at level crossings (T1) .647 **
* p < .05 ** p < .01 *** p < .001
6.5.1.9 Attitudes towards driving at level crossings
The first eight items from the 15-item instrument with bipolar scales (i.e. -3 to +3)
directly measured attitudes towards driving at level crossings. As stated previously, each
of the bipolar scales was recoded using unipolar (1-7) scales. Participants were asked to
complete statements by rating pairs of adjectives whilst driving at level crossings (e.g.
bad-good, more confusing-less confusing). Mean scores indicated that at Time 1 their
was generally a positive attitude towards driving at level crossings, however at Time 2,
two companies (Rogers and Carters) recorded a less positive attitude towards driving at
224
level crossings. Table 32 below illustrates the mean scores of each company at both time
points.
Results of paired sample t-tests indicate that there was no significance difference in
the pre and post-intervention paired data sets for self-reported attitudes towards driving at
level crossings of truck drivers before the intervention (M=4.67, S.D.=1.571) to after the
intervention (M=4.81, S.D.=1.457), t(9) = -0.353, p = 0.733.
Table 32: Attitudes towards driving at level crossings Time 1 Time 2
Company n Mean n Mean Abbotts 12 5.33 5 5.25 Thompsons 6 5.73 1 5.38 Rogers 13 5.12 2 3.44 Carters 13 5.11 1 4.75
1=Negative attitude to 7=Positive attitude
6.5.1.10 Perceived behaviour control whilst driving at level crossings
The second seven items from the 15-item instrument with bipolar scales (i.e. -3 to
+3) directly measured perceived behavioural control (self-efficacy) whilst driving at level
crossings. As stated previously, each of the bipolar scales was recoded using unipolar (1-
7) scales. Participants were asked to complete statements by rating pairs of adjectives
whilst driving at level crossings (e.g. not up to me-up to me, out of my control-under my
control). Mean scores at Time 1 indicated that generally participants believe that they
have control over their driving behaviour at level crossings, however mean scores for
three of the companies dropped at Time 2.
Table 33: Perceived behavioural control whilst driving at crossings Time 1 Time 2
Company n Mean n Mean Abbotts 13 5.58 5 5.60 Thompsons 6 6.05 1 5.57 Rogers 13 5.32 2 4.71 Carters 13 6.03 1 7.00
1=Negative attitude to 7=Positive attitude
225
Results of paired sample t-tests indicate that there was no significance difference in
the pre and post-intervention paired data sets for self-reported self-efficacy whilst driving
at level crossings of truck drivers before the intervention (M=5.14, S.D.=1.467) to after
the intervention (M=5.56, S.D.=0.777), t(9) = -1.010, p = 0.342.
6.5.1.11 Subjective norms of others
Participants were asked a range of questions of their beliefs about how their
colleagues, friends and other motorists drive at level crossings. The means scores
relating to the safe driving behaviour of other motorists was substantially worse than
participant’s colleagues or their friends. This is illustrated with the mean score of ‘other
motorists generally obey the rules at level crossings’ (T1=2.68, T2=3.09) being quite a lot
higher than ‘your colleagues generally obey the rules at level crossings’ (T1=1.55,
T2=1.64) and ‘your friends generally obey the rules at level crossings’ (T1=1.55,
T2=1.82). Mean scores for ‘it is generally possible to judge a train’s speed’ (T1=4.44,
T2=4.00) indicated that participants believe that it is generally not possible to judge a
trains speed. However, the mean score for ‘it is generally safe to disobey the rules at
level crossings’ (T1=3.57, T2=3.91) indicated that some participants believe that they can
disobey the rules at level crossings and not be at risk of a collision with a train. Table 35
below displays the means scores for each of the items in this instrument at Time 1 and
Time 2.
Results of paired sample t-tests indicate that there was no significance difference in
the pre and post-intervention paired data sets for the subjective norms of others (i.e.
friends, colleagues and other motorists) before (M=2.70, S.D.=0.500) or after the
intervention (M=2.80, S.D.=0.469), t(7) = -0.498, p = 0.636.
226
Table 34: Mean scores for subjective norms of others Time 1 Time 2
Item n Mean n Mean Your colleagues generally obey the rules at level crossings 42 1.55 11 1.64
Your friends generally obey the rules at level crossings 42 1.55 11 1.82
Other motorists generally obey the rules at level crossings 40 2.68 11 3.09
Your colleagues generally think it important to obey the rules at level crossings 41 1.41 11 1.64
Your friends generally think it important to obey the rules at level crossings 41 1.46 11 1.64
Other motorists generally think it important to obey the rules at level crossings 42 2.45 11 2.45
It is generally safe to disobey the rules at level crossings 42 3.57 11 3.91
It is generally possible to judge a train’s speed 41 4.44 11 4.00It is generally safe to cross if you can’t see a train, even if the lights are flashing 42 4.10 11 4.27
It is generally safe to roll slowly through a crossing instead of stopping 42 4.02 10 3.80
Trains generally run to a regular timetable 42 3.45 10 4.20Penalties need to be tougher for violating road rules at level crossings 42 2.17 10 1.70
The main deterrent for breaking the rules at level crossings is fear of getting caught 41 3.41 11 2.64
Generally it is more important to use common sense at level crossings than strictly follow the road rules 42 3.31 10 2.90
1=Strongly agree to 5=Strongly disagree
6.5.1.12 Perceived risk of a level crossing collision
Participants were asked about their perception of how likely it is that they would be
involved in a collision at a level crossing whilst driving a truck. The table below
illustrates that at all time points, the mean scores were extremely low, indicating that
most participants believe that they will never be involved in a level crossing collision.
Table 35: Belief of the likelihood of being involved in a level crossing collision Time 1 Time 2
Belief N Mean n Mean Likelihood of collision with train 45 0.33 11 0.09
0=Not at all likely to 5=Very likely
227
Results of paired sample t-tests indicate that there was no significance difference in
the pre and post-intervention paired data sets of perception of being involved in a level
crossing collision before the intervention (M=0.45, S.D.=1.214) to after the intervention
(M=0.09, S.D.=0.302), t(11) = 0.938, p = 0.371.
6.5.1.13 Environmental constraints whilst driving at level crossings
The items ‘blinding sun makes it difficult to see if the red flashing lights are
activated’ (T1=3.16, T2=2.73), ‘other drivers (such as cars) do stupid things that put you
in a dangerous situation’ (T1=2.98, T2=2.55) and ‘the design of the road is an S bend and
it is difficult to see if a train is approaching or at the railway crossing’ (T1=2.71,
T2=2.82), scored the highest means for both time points. Formative research with truck
drivers had indicated that these were significant issues whilst driving a truck at level
crossings, which supports results from this current sample. Table 36: Beliefs of design and environmental factors at level crossings
Time 1 Time 2
Belief n Mean n Mean
Blinding sun makes it difficult to see if the red flashing lights are activated 45 3.16 11 2.73
The design of the road is an ‘S’ bend and it is difficult to see if a train is approaching or at the railway crossing 45 2.71 11 2.82
The height of the truck’s cabin makes it difficult to see a train or the warning systems 45 1.53 11 1.55
Warning systems on the road approaching the crossing are not adequate to inform trucks there is a railway crossing ahead 45 2.49 11 2.00
The mass of the truck makes it difficult to brake in time to stop at a railway crossing 45 1.96 11 2.18
Road surface is often poor and it is difficult to stop 44 2.30 11 2.27Boom gates and/or flashing lights are often faulty 45 2.09 11 1.91Other drivers (such as cars) do stupid things that put you in a dangerous situation 45 2.98 11 2.55
Intersections ahead of a railway crossing often cause your truck to overhang the tracks 45 1.84 11 1.36
Noise from the truck’s engine is too loud to hear an approaching train 45 1.80 11 1.73
When you have to stop at a railway crossing, it takes a long time before your truck is able to get over the crossing 45 2.16 10 2.40
All ‘Stop’ signs should be changed to ‘Give Way’ signs at railway crossings as these are easier for trucks to get through the crossing
45 1.80 11 2.00
My truck has stalled on the tracks at a railway crossing 45 1.09 11 1.00(1=never, 2=almost never, 3=sometimes, 4=almost always, 5=always)
228
Not surprisingly, results of paired sample t-tests indicate that there was no
significance difference in the pre and post-intervention paired data sets of beliefs of
participants about the design and environmental factors influencing safety at level
crossings before (M=2.29, S.D.=0.501) or after the intervention (M=1.98, S.D.=0.380),
t(10) = 1.754, p = 0.113. Although participants were provided with information
acknowledging possible environmental/design difficulties they may face when driving a
truck at level crossings, there was very little difference in beliefs regarding their
perceived environmental influences from being exposed to the intervention message.
6.5.1.14 Familiarity with different protection systems
Approximately 20% (n=10) of participants reported never driving through level
crossings with boom gates whilst driving for work, while 40% of drivers reported driving
through them on a daily basis. Driving through passive crossings with flashing lights or
just a stop/give-way sign was less common than driving through boom gates. Over a
third of participants reported that they never drive over any type of passive crossing
whilst driving a truck. Table 37 illustrates participant reported exposure to driving at
level crossings.
Table 37: Exposure to level crossing driving
Boom gates
Flashing Lights only
No lights or boom gate Type n % n % n % Daily 18 40.0 9 20.0 12 26.7Weekly 12 26.7 8 17.8 5 11.1Monthly 5 11.1 9 20.0 8 17.8Yearly 0 0 3 6.7 5 11.1Never 10 22.2 16 35.6 15 33.3
The three items in the questionnaires were then recoded to compare those
participants that either drive over active crossings or passive crossings regularly. Those
participants that stated they never or yearly drove over either active or passive crossings
were combined to form a group of being unfamiliar with that type of crossing whilst
driving a truck. The recoding of these variables is used for the inferential statistics.
229
A series of one-way ANOVA’s was then calculated to examine differences between
familiarity with level crossings (boom gates, flashing lights or passive) in relation to each
of the driving instruments.
Table 38: Mean scores for familiarity at level crossings (boom gates) Instrument Familiarity N df Mean F Sig DBQ Familiar 34 1, 41 0.65 3.284 .770 Unfamiliar 9 0.47 Self-reported driving behaviour Familiar 35 1, 43 1.27 4.918 .320
Unfamiliar 10 1.07 Intended driving behaviour Familiar 34 1, 42 1.21 .005 .946 Unfamiliar 10 1.22 Subjective norms Familiar 34 1, 42 5.89 0.250 .620 Unfamiliar 10 5.71 Attitude Familiar 34 1, 42 5.25 .003 .953 Unfamiliar 10 5.28 Perceived behavioural control Familiar 35 1, 43 5.79 1.499 .227 Unfamiliar 10 5.39 Beliefs about environmental constraints Familiar 34 1, 42 2.17 0.683 .413
Unfamiliar 10 2.02
There were no significant results in relation to drivers that are familiar with boom
gated crossings and any of the driving instruments.
Table 39: Mean scores for familiarity at level crossings (flashing lights)
Instrument Familiarity N df Mean F Sig DBQ Familiar 25 1, 41 0.69 5.534 .024 Unfamiliar 18 0.50 Self-reported driving behaviour Familiar 26 1, 43 1.31 8.591 .005
Unfamiliar 19 1.10 Intended driving behaviour Familiar 25 1, 42 1.26 1.267 .267 Unfamiliar 19 1.14 Subjective norms Familiar 25 1, 42 6.09 3.754 .059 Unfamiliar 19 5.52 Attitude Familiar 19 1, 42 5.40 .829 .368 Unfamiliar 25 5.07 Perceived behavioural control Familiar 26 1, 43 6.03 9.355 .004 Unfamiliar 19 5.25 Beliefs about environmental constraints Familiar 26 1, 42 2.25 3.215 .080
Unfamiliar 18 1.97
230
A significant effect for the driver behaviour questionnaire F (1, 41) = 5.534, p
<0.05, self-reported driving behaviour at level crossings F (1, 43) = 8.591, p <0.05 and
self-efficacy whilst driving at level crossings F (1, 43) = 9.355, p <0.05 were revealed.
This indicated that truck drivers that were familiar with driving at crossings with only
flashing lights recorded higher levels of unsafe driving at level crossings generally,
slightly higher risk taking driving behaviour and higher levels of self-efficacy/perceived
behavioural control than unfamiliar drivers.
Table 40: Mean scores for familiarity at level crossings (passive signs only)
Instrument Familiarity n df Mean F Sig DBQ Familiar 24 1, 41 0.66 1.978 .167 Unfamiliar 19 0.55 Self-reported driving behaviour Familiar 25 1, 43 1.27 1.567 .217
Unfamiliar 20 1.17 Intended driving behaviour Familiar 24 1, 42 1.20 0.013 .909 Unfamiliar 20 1.22 Subjective norms Familiar 24 1, 42 5.89 0.093 .762 Unfamiliar 20 5.80 Attitude Familiar 24 1, 42 5.26 0.000 .998 Unfamiliar 20 5.26 Perceived behavioural control Familiar 25 1, 43 5.79 0.533 .469 Unfamiliar 20 5.59 Beliefs about environmental constraints Familiar 25 1, 42 2.27 3.867 .056
Unfamiliar 19 1.96
There were no significant results in relation to drivers that are familiar with passive
crossings (stop or give-way sign only) and any of the driving instruments.
6.5.1.15 Intervention message recall
All participants received an intervention message immediately after completion of
the first questionnaire (Time 1). The second questionnaire (Time 2) asked participants if
they recalled the radio message they received about level crossings, what slogans or
information they recalled (if any) and how likely it was that their driving behaviour was
influenced from receiving the message. At Time 2 (post-test), five participants recalled
having received the intervention message about level crossings, while four participants
231
stated that they received a message about fatigue. As there was no intervention message
about fatigue, these four participants obviously didn’t recall the intervention at all.
However, no participant could recall any slogans or information provided to them in the
intervention about level crossings.
6.5.2 Older drivers
6.5.2.1 Sample attrition
At Time 1, 152 participants completed questionnaires, with this figure dropping to
109 at Time 2. The attrition rates for intervention and control groups are displayed in the
table below. Additionally, this table illustrates the attrition for metropolitan and
rural/remote participants. Interestingly, the number of participants in intervention and
control groups remained equal at both time points. With regards to area classification, at
Time 1 there was similar percentages of metropolitan (55.3%) and rural/remote (44.1%)
participants.
Table 41: Participant group and area classification at both time points Time 1 Time 2
n % n % Group Intervention 76 50.0 54 49.5 Control 76 50.0 55 50.5 Total 152 - 109 - Area Classification Metropolitan 84 55.3 70 64.2 Rural/remote 67 44.1 39 35.8 Missing 2 0.7 0 0.0
6.5.2.2 Demographics
At Time 1, the mean age for participants was 69.2 years (S.D. 6.842), with the
youngest participant being 60 years and the oldest participant being 89 years. At Time 2
the mean age of participants was approximately the same (69.4 years). At Time 1, more
than two-thirds of participants were male (69.1%) and 21.7% of participants reported still
232
being in paid work. Approximately half of all participants reported having a condition on
their drivers licence as described in Table 42 below.
Table 42: Licence conditions as reported by participants Time 1 Condition n % No condition 77 50.7 S condition only (corrective lenses required) 46 30.3 M condition only (medical certificate require) 12 7.9 Both S & M conditions 16 10.5 Missing 1 0.7
Older drivers were also asked about their any medical conditions that they suffer
from. Nearly half of the sample (42.2%) reported high blood pressure and a fifth (22.4%)
reported either some hearing loss or deafness. Twenty-seven participants (17.8%)
reported either neck, back or limb disorders. With safe driving at level crossings
involving both listening for trains and movement of the neck to scan for trains, it is of
concern that so many participants in this sample reported medical conditions that restrict
both of these behaviours. This may be one of the factors that is important in the over-
representation of older drivers in level crossing collisions.
233
Table 43: Medical conditions suffered by participants Time 1 Medical Condition n % High blood pressure 64 42.1 Heart disease 16 10.5 Chest pain/Angina 19 12.5 Any condition requiring heart surgery 11 7.2 Palpitations/Irregular heart beat 14 9.2 Head/Spinal injury 8 5.3 Seizures, fits, convulsions or epilepsy 1 0.7 Abnormal shortness of breath 4 2.6 Blackouts or fainting 2 1.3 Stroke 2 1.3 Dizziness or vertigo 10 6.6 Double vision 3 2.0 Colour blindness 1 0.7 Kidney disease 3 2.0 Diabetes 10 6.6 Neck, back or limb disorders 27 17.8 Hearing loss or deafness 34 22.4
6.5.2.3 Self-assessment of driving ability
Participants were asked a range of questions relating to their driving ability at Time
1. All mean scores for driving ability were relatively good (i.e. close to 0) indicating that
participants in general believe that have no problems with driving. The highest mean
observed was ‘having more trouble adjusting to glare and/or night driving than you did
previously’ (1.34), while the mean for ‘have regular health and vision checks’ was 3.82.
It appears that vision is of the greatest driving concern from these items, yet not all
drivers are having regular check-ups. Table 44 below illustrates the means for each of
the items in this instrument.
234
Table 44: Driving ability Item n Mean Reactions to unexpected situations slower than they used to be 151 1.06Have trouble judging the distance of other vehicles, or changing focus from your instrument panel to the road 152 0.27
Having more trouble adjusting to glare and/or night driving than you did previously 152 1.34Ever get surprised by pedestrians or other vehicles coming from your left or right while you are focusing straight ahead 152 0.55
Some traffic situations or other drivers upset you 151 1.21Have trouble driving through, or turning at busy intersections or roundabouts 152 0.36Feel uncomfortable driving in unfamiliar territory 152 1.06Find that you are easily distracted or that your thoughts wander while you are driving 152 0.53Have regular health and vision checks 152 3.82
0=Not at all to 5=Very often
Health and vision checks was then recoded to be aligned with the other items (i.e.
‘Very often’ was recoded to 0 and ‘Not at all’ was recoded to 5). This ensured that those
participants that reported having regular checks would receive a low score for this
question when the total instrument score was calculated. The total instrument score for
each participant was then recoded into three categories: green rating (0-5), yellow rating
(6-15) and red rating (16+). These categories represented the self-reported driving ability
of participants (green=no problems driving, yellow=some problems driving, red=quite a
lot of problems driving). The purpose of recoding into three categories was to compare
ratings between gender and rurality. Table 45 below demonstrates the differences
between gender and rurality. Percentages of ratings are shown as being within the group
(i.e. gender or rurality).
Table 45: Driving self-assessment ratings between groups Green Rating Yellow Rating Red Rating Total
Participants in Group
(n) n % n % n %
Gender Male 105 39 37.5 59 56.7 6 5.8 Female 47 12 26.1 31 67.4 3 6.5 Rurality Metropolitan 84 28 33.7 51 61.4 4 4.8 Rural/remote 67 23 34.8 39 59.1 4 6.1
235
From Table 45 it is clear that a greater percentage of male participants report
having very few problems driving (green rating =37.5%) compared to female participants
(green rating=26.1%). There was little difference in ratings between metropolitan and
rural/remote participants. Analysis of variances was also undertaken to examine
differences between gender and rurality and the mean scores for this instrument. No
significant effect was found for gender and self-reported driving ability F (1, 149) =
0.974, p =.325 or rurality and self-reported driving ability F (2, 149) = 2.241, p =.110.
6.5.2.4 Crash history
At Time 1, 29 participants in total reported having been involved in a road crash
during the past 3 years. Table 46 below illustrates the differences between rural/remote
and metropolitan driver crash involvement.
Table 46: Comparison of crash involvement and rurality at Time 1 Rurality
Involved in a crash
(n)
Involved in a crash (%)
Total participants
(n) Metropolitan 20 23.8 84 Rural/remote 9 13.4 67 Missing - - 1 Total 29 19.1 152
Metropolitan drivers (23.8%) were more likely to have been involved in a collision
during the past 3 years than rural/remote drivers (13.4%), which is not surprising given
the larger driver population in metropolitan areas. Five participants reported having been
involved in 2 collisions, while one participant reported having been involved in three
collisions during the past 3 years. One metropolitan participant reported having been
involved in a level crossing collision with a fatality occurring in that collision. The
majority of drivers that reported having been involved in a collision during past 3 years
indicated that it was damage only (89.7%) and an angular/side on collision (44.8%).
236
Table 47: Road crashes during past 3 years Time 1 n % * Seriousness Damage only 26 89.7 Slight injury (to any person) 1 3.4 Serious injury (to any person) 0 0 Fatality (to any person) 1 3.4 Level crossing collision 1 3.4 Type Head on collision with another vehicle 0 0 Rear-end collision with another vehicle 10 34.5 Angular (i.e. side-on) collision with another vehicle 13 44.8 Collision with a pedestrian 0 0 Collision with another object (e.g. parked car, animal, tree) 5 17.2 Overturned vehicle 0 0
* Percentage of drivers that reported having been involved in a crash during past 3 years (i.e. 29 drivers in total)
Self-reported contributing factors to these collisions revealed that five drivers
(17.2%) were distracted by something outside the vehicle, two were daydreaming (6.8%)
and two were listening to the radio (6.8%). Table 48 below illustrates the self-reported
contributing factors to road crashes during the past 3 years.
Table 48: Self-report contributing factors to road crash Time 1 Type n %* Been drinking alcohol before driving 0 0 Felt tired 1 3.4 Driving too fast for the conditions 0 0 Talking to a passenger in the vehicle 1 3.4 Listening to the radio 2 6.8 Talking on a mobile phone 0 0 Trying to pick up something from the seat or floor 0 0 Adjusting the radio/cassette/CD or fan/air conditioning 0 0 Checking the instruments (e.g. fuel gauge or speedometer) 0 0 Daydreaming 2 6.8 Distracted by something inside the vehicle 1 3.4 Distracted by something outside the vehicle 5 17.2
* Percentage of the 29 drivers that reported being involved in a crash during past 3 years.
237
Analysis of variances was also undertaken to examine differences between
participants that reported being involved in a road crash during the past 3 years and each
of the driving instruments at Time 1.
Table 49: Mean overall scores between road crashes and driving instruments
Instrument Crash History n df Mean F Sig DBQ Crash 28 1, 142 0.42 .008 .931 No crash 115 0.43 Self-reported driving behaviour Crash 29 1, 140 0.10 .512 .476
No crash 112 0.13 Intended driving behaviour Crash 26 1, 137 0.11 .707 .402 No crash 112 0.14 Subjective norms Crash 26 1, 136 3.80 2.459 .119 No crash 111 3.62 Attitude Crash 26 1, 140 5.57 .334 .564 No crash 115 5.43 Perceived behavioural control Crash 26 1, 142 6.04 3.047 .083 No crash 117 5.76 Beliefs about environmental constraints Crash 28 1, 140 1.52 4.424 .037*
No crash 113 1.77
A significant effect was found for one of the driving instruments. An effect was
revealed for beliefs about environmental constraints when driving at level crossings F (1,
140) = 4.424, p <0.05. Therefore, this indicates that older drivers that reported having
had a crash during the past 3 years were less likely to believe that environmental
constraints influences their ability to drive safely at level crossings. Therefore, drivers
who hadn’t had a crash during the past 3 years were more likely to believe that
environmental factors play a role in safe driving behaviour at level crossings.
6.5.2.5 Knowledge of level crossing road rules
Participants were asked four questions about road rules and facts about level
crossings. Only a fifth of drivers (19.1%) answered all questions correctly, while four
participants (2.6%) did not answer any questions correctly. This is an interesting finding
as anecdotal evidence suggested that drivers have a poor knowledge of the road rules at
level crossings but it had never been confirmed by research in Australia.
238
6.5.2.6 General driver behaviour
All mean scores for items on the modified driver behaviour questionnaire (DBQ)
were very low (i.e. less than 1) at both time points. The mean sample score at Time 1
was 0.43 (S.D.=0.266) which remained relatively similar at Time 2 (M=0.45,
S.D.=0.290).
6.5.2.7 Self-reported driving behaviour at level crossings
At Time 1, the mean score for this instrument was 0.12 (S.D.=0.183) indicating that
participants believe they drive safely at level crossings.
6.5.2.8 Intended driving behaviour at level crossings
Sample mean scores for intended driving behaviour at level crossings at Time 1
(M=0.14, S.D.=0.202) were very low, indicating that participants intend to drive safely at
level crossings in the future. A correlation analysis was undertaken to explore the
relationships between the dependent variable driving intention at level crossings (Time 1)
and the independent variables. Table 50 illustrates the significant relationships that were
revealed. As shown in this table, there was a weak correlation between participants’ self-
reported driving ability and driving intention at level crossings (Time 1) [r = .196, p <
.05]. In other words, a self-reported poorer driving ability was associated with a stronger
intention to perform high risk behaviours at level crossings. As expected, driving
intention at level crossings (Time 1) and self-reported driving behaviour at level
crossings (Time 1) were highly correlated [r = .792, p < .01].
239
Table 50: Bivariate correlations for older drivers Dependent Variable
Independent variables Driving Intention at Level Crossings (T1)
Age .038 Gender -.050 Conditions on licence .077 Rurality .048 Hours per week drive -.132 Driving ability (self-reported) .196 * Crashes during past 3 years .072 DBQ .119 Attitude -.015 Perceived behavioural control .004 Subjective norms -.132 Environmental constraints .167 Perceived risk of collision -.042 Driving intention - Self-reported driving behaviour .762 **
* p < .05 ** p < .01 *** p < .001
To further analyse the prediction of driving intention at level crossings, a
hierarchical multiple regression was used. This 3-step hierarchical regression analysis
was undertaken to assess the contribution of the components of the integrated model (IM)
to the prediction of behavioural intention, along with measures of environment,
skills/ability (self-reported driving ability) and perceived risk. Three blocks of variables
were used to predict intention at Time 1: (i) environmental constraints and driving ability,
(ii) attitudes, subjective norms and perceived behavioural control, and (iii) perceived risk.
In this way, it was possible to assess the additional predictive utility of the attitudes,
subjective norms, and perceived behaviour control and the ability of environmental
constraints and compliance with road laws to mediate the influence of perceived risk.
The overall 3-step regression model was not significant, and no outcome variables
were found to be predictors of intention to drive safely at level crossings. At Step 1,
environmental constraints and driving ability variables accounted for only 4.4% of the
variance in behavioural intention. The addition of attitudes, subjective norms and
perceived behavioural control (F change = .771, NS) led to small (6.3%) increases in the
amount of variance explained, however the final addition of perceived risk (F change =
240
1.178, NS) led to even smaller (7.3%) increases in the overall amount of variance
explained by the model.
Table 51: Hierarchical regression of constructs on intention at level crossings Variables Mean S.D. B Std
Error β R2 Adj R2 ∆ R2
Step 1 Environmental constraints 1.73 .567 .055 .035 .153
Driving ability .89 .468 .071 .041 .163 .044 .028 Step 2 Attitude 5.46 1.152 -.001 .018 -.004 Subjective norms 3.07 .449 -.055 .043 -.121 Perceived behavioural control 5.81 .728 -.028 .028 -.101
.063 .022 Step 3 Perceived risk .21 .721 -.029 .026 -.101 .073 .024 .010
* p<.05 ** p<.01 ***p<.001
6.5.2.9 Attitudes towards driving at level crossings
The first eight items from the 15-item instrument with bipolar scales (i.e. -3 to +3)
directly measured attitudes towards driving at level crossings. As stated previously, each
of the bipolar scales was recoded using unipolar (1-7) scales. Participants were asked to
complete statements by rating pairs of adjectives whilst driving at level crossings (e.g.
bad-good, more confusing-less confusing). Mean scores indicated that at Time 1 their
was generally a positive attitude towards driving at level crossings, however one item
relating to road rules suggested that participants were unsure if they were not strict
enough or too strict (M=3.85).
241
Table 52: Attitudes towards driving at level crossings Time 1
Item n Mean Design of crossings (bad) 146 5.18Design of crossings (unsafe) 148 5.02Design of crossings (confusing) 150 5.85Design of crossings (difficult to obey rules) 147 5.91Road rules (bad) 147 6.12Road rules (not strict enough) 147 3.85Road rules (confusing) 150 5.79Road rules (not practical) 146 5.94
1=Negative attitude to 7=Positive attitude
6.5.2.10 Perceived behavioural control whilst driving at level crossings
The second seven items from the 15-item instrument with bipolar scales (i.e. -3 to
+3) directly measured self-efficacy and perceived behaviour control whilst driving at
level crossings. As stated previously, each of the bipolar scales was recoded using
unipolar (1-7) scales. Participants were asked to complete statements by rating pairs of
adjectives whilst driving at level crossings (e.g. not up to me-up to me, out of my control-
under my control). Mean scores at Time 1 indicated that participants generally believe
that they have control over their driving at level crossings, however one item ‘other
motorists influence my driving’ (harder to obey rules to easier to obey rules) scored a
lower mean (M=4.95) which indicated that some participants feel that they can’t always
obey the road rules at level crossings because of other motorists behaviour. Such
behaviour may include peer pressure for older drivers. Table 53 below illustrates the
mean scores observed at Time 1 for perceived behavioural control (self-efficacy).
Table 53: Perceived behavioural control whilst driving at level crossings
Time 1 Item n Mean Obeying the rules (not up to me) 149 6.75Obeying the rules (out of my control) 147 6.49Obeying the rules (dependent on other motorists) 148 6.55Obeying the rules (dependent on time constraints) 148 6.65Other motorists influence my driving (harder to obey rules) 146 4.95Other motorists influence my driving (more confusing) 144 5.45Other motorists influence my driving (more stressful) 144 5.20
1=Negative attitude to 7=Positive attitude
242
6.5.2.11 Subjective norms of important others
Older drivers were asked a range of questions about how their family, friends and
other motorists drive at level crossings. Mean scores for family and friends obeying the
rules at level crossings were substantially higher than other motorists obeying the rules at
level crossings at Time 1. This indicated that older participants believe that motorists
other than their family and friends do not obey the rules at level crossings as often as their
important others (i.e. family and friends). The item ‘penalties need to be tougher for
violating road rules at crossings’ scored a relatively high mean indicating that older
participants believe that the current penalties are not tough enough for violators. Table
54 below illustrates the mean scores of each item at Time 1.
Table 54: Subjective norms of others Time 1
Item n Mean Your family generally obeys the rules at rail crossings 149 1.21Your friends generally obey the rules at rail crossings 143 1.66Other motorists generally obey the rules at rail crossings 147 2.71Your family generally think it important to obey the rules at rail crossings 150 1.17Your friends generally think it important to obey the rules at rail crossings 142 1.56Other motorists generally think it important to obey the rules at rail crossings 146 2.50
It is generally safe to disobey the rules at rail crossings 151 4.69It is generally possible to judge a train’s speed 150 4.35It is generally safe to cross if you can’t see a train, even if the lights are flashing 151 4.74
It is generally safe to roll slowly through a crossing instead of stopping 151 4.60Trains generally run to a regular timetable 149 3.81Penalties need to be tougher for violating road rules at rail crossings 149 2.05The main deterrent for breaking the rules at rail crossings is fear of getting caught 147 3.30
Generally it is more important to use common sense at rail crossings than strictly follow the road rules 151 3.87
1=Strongly agree to 6=Strongly disagree
6.5.2.12 Perceived risk of a level crossing collision
Participants were asked about their perception of how likely it is that they would be
involved in a collision at a level crossing whilst driving. At Time 1, the perceived risk
was extremely low (M=0.19).
243
6.5.2.13 Environmental constraints whilst driving at level crossings
The items ‘my car stalled on the tracks at a rail crossing’ and ‘intersections ahead
of a rail crossing often cause your car to overhang the tracks’ scored the lowest mean at
Time 1. At Time 1 (M=2.64) the item ‘difficult to hear an approaching train when the
windows are up’ scored the highest mean. With such relatively low means across all
items, it indicates that older participants don’t believe that the design and environment
play a crucial role in their driving at level crossings. Table 55 illustrates the mean scores
for all items at Time 1.
Table 55: Beliefs of environmental factors at level crossings
Time 1 Item n Mean Blinding sun makes it difficult to see if the red flashing lights are activated 146 2.00The design of the road makes it difficult to see if a train is approaching or at the rail crossing 147 1.90
Warning systems on the road approaching the crossing are not adequate to inform drivers there is a rail crossing ahead 145 2.04
Road surfaces are often poor and it is difficult to stop 146 1.50Boom gates and/or flashing lights are often faulty 145 1.40Other drivers do stupid things that put you in a dangerous situation 145 1.81Intersections ahead of a rail crossing often cause your car to overhang the tracks 145 1.20
Difficult to hear an approaching train when the windows are up 145 2.64My car stalled on the tracks at a rail crossing 147 1.00
1=Never, 2=Almost never, 3=Sometimes, 4=Almost always, 5=Always
6.5.2.14 Familiarity with difference protection systems
Table 56 illustrates participant reported exposure to driving at level crossings. The
majority of participants reported that they never drive through boom gates (57.2%), never
drive through crossings with flashing lights (61.2%) and never drive through passive
crossings without lights or boom gates (59.9%).
244
Table 56: Exposure to level crossing driving Boom gates Flashing Lights only No lights or boom gate
Type n % n % n % Never 87 57.2 93 61.2 91 59.9Once a year 9 5.9 14 9.2 20 13.2Twice a year 14 9.2 15 9.9 8 5.3Monthly 14 9.2 11 7.2 6 3.9Weekly 17 11.2 11 7.2 13 8.6Daily 8 5.3 4 2.6 9 5.9Missing 3 2.0 4 2.6 5 3.3
The exposure instrument was then recoded to compare those participants that either
drive through each type of crossing regularly or not. Those participants that stated they
never, yearly or twice a year drove over either active or passive crossings were combined
to form a group of being unfamiliar with that type of crossing whilst driving. Not
surprisingly participants that lived in rural or remote areas were more familiar with
crossings that had only a stop or give-way sign and red flashing lights, however, the same
percentage of metropolitan and rural/remote participants reported being unfamiliar with
boom gates (i.e. 74%).
Table 57: Familiarity and area classification at Time 1 Familiar Unfamiliar
Protection System n % n % Boom gates Metropolitan 21 13.8 61 40.1 Rural/remote 17 11.2 49 32.2 Flashing red lights Metropolitan 6 3.9 75 49.3 Rural/remote 20 13.2 46 30.3 Give-way or Stop sign only Metropolitan 9 5.9 72 47.3 Rural/remote 19 12.5 46 3.9
A series of one-way analysis of variances were calculated to examine differences
between familiarity of older drivers with level crossings (boom gates, flashing lights or
passive) in relation to each of the driving instruments. Drivers that were considered
245
‘familiar’ with level crossings were those that reported driving over a type of crossing
daily, weekly or monthly. Results are presented in the tables below.
Table 58: Familiarity with driving at level crossings (boom gates) Instrument Familiarity n df Mean F Sig DBQ Familiar 37 2, 138 0.52 3.986 .021* Unfamiliar 103 0.39 Self-reported driving behaviour Familiar 39 2, 136 0.17 2.180 .117 Unfamiliar 99 0.10 Intended driving behaviour Familiar 36 1, 135 0.17 1.686 .196 Unfamiliar 100 0.12 Subjective norms Familiar 36 1, 135 3.58 0.969 .327 Unfamiliar 100 3.68 Attitude Familiar 36 1, 138 5.26 1.515 .221 Unfamiliar 103 5.52 Perceived behavioural control Familiar 38 1, 140 5.86 0.231 .631 Unfamiliar 103 5.79 Beliefs about environmental constraints Familiar 38 1, 138 1.69 0.118 .732
Unfamiliar 101 1.73 * p<.05
The instrument ‘Driver Behaviour Questionnaire’ was the only significant factor for
familiarity with boom gated crossings [F (2, 138) = 3.986, p <0.05]. This indicated that
older drivers that were familiar with driving at crossings with boom gates recorded higher
levels of unsafe driving generally than unfamiliar drivers.
246
Table 59: Familiarity with driving at level crossings (flashing lights) Instrument Familiarity n df Mean F Sig DBQ Familiar 26 1, 139 0.39 .584 .446 Unfamiliar 114 0.44 Self-reported driving behaviour Familiar 26 1, 135 0.21 7.989 .005*
Unfamiliar 111 0.10 Intended driving behaviour Familiar 25 1, 135 0.22 5.707 .018* Unfamiliar 111 0.12 Subjective norms Familiar 23 1, 133 3.62 .049 .825 Unfamiliar 111 3.65 Attitude Familiar 26 1, 138 5.23 1.336 .250 Unfamiliar 113 5.51 Perceived behavioural control Familiar 26 1, 140 5.77 .119 .730 Unfamiliar 115 5.82 Beliefs about environmental constraints Familiar 26 1, 138 1.73 .006 .939
Unfamiliar 113 1.72 * p<.05
A significant effect for two instruments: self-reported driving behaviour F (1, 135)
= 7.989, p <0.05 and intended driving behaviour F (1, 135) = 5.707, p <0.05 were
revealed. This indicated that older drivers that were familiar with driving at crossings
with only flashing lights recorded higher levels of unsafe driving and intended driving
behaviour at level crossings than unfamiliar drivers.
Table 60: Familiarity of driving at level crossings (passive signs only) Instrument Familiarity n df Mean F Sig DBQ Familiar 26 1, 139 0.38 .714 .399 Unfamiliar 114 0.43 Self-reported driving behaviour Familiar 28 1, 136 0.19 3.763 .054
Unfamiliar 109 0.11 Intended driving behaviour Familiar 28 1, 135 0.16 .390 .533 Unfamiliar 108 0.13 Subjective norms Familiar 26 1, 133 3.62 .101 .751 Unfamiliar 108 3.66 Attitude Familiar 27 1, 138 5.07 4.343 .039* Unfamiliar 112 5.56 Perceived behavioural control Familiar 27 1, 140 5.77 .174 .677 Unfamiliar 114 5.83 Beliefs about environmental constraints Familiar 27 1, 1.77 .185 .668
Unfamiliar 111 137 1.72 * p<.05
247
A significant result was found for the scale attitudes towards driving at level
crossings F (1, 138) = 4.343, p <0.05 and familiarity with passive crossings (stop or give-
way sign only). This indicated that participants that are familiar with driving at passive
crossings are more likely to have a negative attitude towards driving at level crossings
generally.
6.5.2.15 Intervention message recall
All participants received an intervention message via telephone six weeks after
returning their Time 1 questionnaire. Of the 76 control participants, only 33 indicated
that they remembered what radio message they received (43.4%). Forty-four (44)
participants in the intervention group (n=76) recalled that they had received a radio
message about level crossing safety (57.9%). These low recall figures were staggering
given that all participants were prompted of either receiving a message about vision or
level crossings. Of the participants that recalled receiving the radio messages, only nine
reported remembering any slogans while 23 reported remembering information from the
radio messages (14 control participants and 9 intervention participants). Participants were
also asked how likely it was that their driving behaviour was influence from receiving
either the control or intervention message. Only two intervention and three control
participants indicated that it was ‘very likely’ that their driving behaviour had been
affected by listening to the message. Therefore, it can be concluded that there generally
was very poor recall of both the intervention and control messages.
6.5.2.16 Effect of intervention exposure
A comparison of means between the intervention and control groups was
undertaken at both Time 1 and Time 2. Table 61 illustrates the mean scores for all
instruments at Time 1 and Time 2 for both the intervention and control groups for the
older driver sample.
248
Table 61: Comparison of experimental group outcome variables Intervention Group Control Group
Instrument
T1 Mean
T2 Mean
T1 Mean
T2 Mean
DBQ .41 .45 .44 .45Self-reported driving behaviour .13 .08 .12 .09Intended driving behaviour .12 .11 .15 .14Attitudes 5.61 5.47 5.32 5.45Perceived behavioural control 5.86 5.94 5.76 5.77Subjective norms 3.06 3.18 3.00 3.05Beliefs of environmental constraints 1.62 1.70 1.77 1.61Perceived risk of a crash .19 .19 .20 .53
To ascertain the effect of the intervention, a split-plot in time analysis (repeated
measures) was undertaken for each of the outcome variables. Separate analyses were
undertaken rather than multivariate ANOVAs since the dependent variables were
predicted to have different outcomes and the outcomes of interest were the individual
variables rather than the set of variables. Results of the series of split-plot analyses reveal
that beliefs about environmental constraints (F = 4.769, df = 1, 95, P< 0.05) was the
only outcome variable to have significantly changed over time between groups. The
associated Partial Eta squared was .048, indicating a small effect size. Perceived risk of a
collision at a level crossing (F = 4.651, df = 1, 104, P< 0.05) was observed to have a
significant time effect (within-subject test), however between groups tests indicates that
the variable group is not significant. Results from the repeated measure analyses are
presented in Table 62.
249
Table 62: Repeated measure analysis of variance for the outcome variables Instrument df Wilks’ A
Effect sizes(Partial Eta
squared)F Sig
DBQ
Time 1, 96 .972 .028 2.737 .101
Group x Time 1, 96 .977 .023 2.241 .138
Self-reported driving behaviour
Time 1, 94 .984 .016 1.486 .226
Group x Time 1, 94 .999 .001 .117 .733
Intended driving behaviour
Time 1, 100 .994 .006 .611 .436
Group x Time 1, 100 .984 .016 1.598 .209
Attitude
Time 1, 100 .994 .006 .611 .436
Group x Time 1, 100 .984 .016 1.598 .209
Perceived behavioural control
Time 1, 100 .985 .015 1.547 .217
Group x Time 1, 100 .998 .002 .221 .639
Subjective norms
Time 1, 87 .977 .023 2.050 .156
Group x Time 1, 87 1.000 .000 .004 .950
Environmental constraints
Time 1, 95 .994 .006 .581 .448
Group x Time 1, 95 .952 .048 4.769 .031*
Perceived risk of a collision Time 1, 104 .957 .043 4.651 .033*
Group x Time 1, 104 .979 .021 2.279 .134
* p<.05 ** p<.01 ***p<.001 Note: to interpret effect sizes: .02=small magnitude; .15=medium magnitude; .35=large magnitude (Cohen, 1988).
250
6.5.3 Younger drivers
6.5.3.1 Sample attrition
At Time 1, 149 participants completed the online questionnaire, with this figure
dropping to 88 at Time 2. The attrition rates for intervention and control groups are
displayed in the table below. Interestingly, like the older driver group, the ratio for
intervention to control remained relatively similar between time points. Additionally, this
table illustrates the attrition for metropolitan and rural/remote participants. As can be
seen, the majority of participants at both time points were from metropolitan areas.
Table 63: Participant group and area classification at both time points Time 1 Time 2
n % n % Group Intervention 100 67.1 58 65.9 Control 49 32.9 30 34.1 Total 149 100.0 88 100.0 Area Classification Metropolitan 92 61.7 53 60.2 Rural/remote 57 38.3 35 39.8 Total 149 100.0 88 100.0
6.5.3.2 Demographics
At Time 1, the majority of participants were female (n=83, 55.7%), had an open
licence (n=118, 79.2%) and were employed (n=135, 90.6%). The mean age of
participants was 21.64 years (range 18 to 24 years). The education level attained by
participants is illustrated in Table 64. The majority of participants reported having either
received a university degree or having done some university (49.0%).
251
Table 64: Education level of younger participants Time 1 Level n % Some high school 7 4.7 Year 12 40 26.8 Some Uni 28 18.8 Uni Degree 45 30.2 Some TAFE 3 2.0 TAFE Degree 19 12.8 Other 7 4.7
To examine the differences between licence type and each of the driving
instruments, a series of analysis of variances was undertaken. The influence of gender on
driving was also examined using a series of ANOVA’s. Tables 65 & 66 illustrate the
ANOVA’s undertaken.
Table 65: Mean overall scores between licence type and outcome variables
Instrument Licence n df Mean F Sig DBQ Provisional 31 1, 148 0.60 .012 .913 Open 118 0.61 Self-reported driving behaviour Provisional 31 1, 148 0.31 .226 .635
Open 118 0.36 Intended driving behaviour Provisional 31 1, 148 0.26 .702 .404 Open 118 0.35 Subjective norms* Provisional 31 1, 148 3.55 .741 .846 Open 118 3.32 Attitude Provisional 31 1, 148 4.06 .269 .605 Open 118 3.97 Perceived behavioural control Provisional 31 1, 148 4.98 .583 .446 Open 118 5.12 Beliefs about environmental constraints Provisional 31 1, 148 2.25 2.184 .142
Open 118 2.50 * Instrument excluded first 3 items due to large amounts of missing data
There were no significant results in relation to licence type and any of the driving
instruments.
252
Table 66: Mean overall scores between gender and outcome variables Instrument Licence n df Mean F Sig DBQ Male 66 1, 148 0.65 1.867 .174 Female 83 0.57 Self-reported driving behaviour Male 66 1, 148 0.46 4.878 .029*
Female 83 0.27 Intended driving behaviour Male 66 1, 148 0.44 4.549 .035* Female 83 0.25 Subjective norms* Male 66 1, 137 3.40 1.148 .291 Female 83 3.31 Attitude Male 66 1, 148 4.03 .246 .620 Female 83 3.96 Perceived behavioural control Male 66 1, 148 4.97 2.272 .134 Female 83 5.19 Beliefs about environmental constraints Male 66 1, 148 2.48 .199 .656
Female 83 2.42 * Instrument excluded first 3 items due to large amounts of missing data
A significant effect was found for two of the driving instruments. An effect was
revealed for self-reported driving behaviour at level crossings [F (1, 148) = 4.878, p
<0.05] as well as intended driving behaviour at level crossings [F (1, 148) = 4.549, p
<0.05]. Therefore, this indicates that young male drivers were significantly more likely
to report current and predict their future driving at level crossings to be less than their
female counterparts. This is not surprising given the amount of literature supporting
research findings that young males are more likely to take risks whilst driving (Laapotti
and Keskinen, 1998, Arnett, 2002).
6.5.3.3 Crash history
Fifty-nine (39.6%) participants reported having been involved in a collision during
the past 3 years. Table 67 below illustrates the differences in crash involvement between
rural/remote and metropolitan participants.
253
Table 67: Comparison of crash involvement and rurality at Time 1 Rurality
Involved in a crash
(n)
Involved in a crash (%)
Total participants
(n) Metropolitan 44 47.8 92 Rural/remote 15 26.3 57 Total 59 39.6 149
Metropolitan drivers (47.8%) were more likely to have been involved in a collision
during the past 3 years than rural/remote drivers (26.3%), which is not surprising given
the larger driver population in metropolitan areas. Five participants reported having been
involved in 2 collisions, while one participant reported having been involved in three
collisions during the past 3 years. One metropolitan participant reported having been
involved in a level crossing collision with a fatality occurring in that collision.
The majority of participants involved in a collision, reported that it was damage
only (79.6%), and 11 participants reported that it was at an intersection (18.6%). Three
participants (5.1%) reported that they were involved in a roll-over, while 28.8% (n=17)
stated that the collision was their fault completely. Participants reported a variety of
contributing factors to their collisions. The most prevalent contributing factor was
reported as ‘listening to music’ (50.8%), followed by ‘daydreaming’ (16.9%), ‘talking to
a passenger’ (15.2%) and ‘speeding’ (13.6%). Ten percent (n=6) of participants reported
fatigue as contributing to their crash, while only one participant reported talking on a
mobile phone as contributing to their crash.
Analysis of variances was also undertaken to examine differences between
participants that reported being involved in a road crash during the past 3 years and each
of the outcome variables at Time 1. No significant effects were found for crash
involvement and any of the driving instruments (see Table 68).
254
Table 68: Mean overall scores between road crashes and outcome variables Instrument Crash History n df Mean F Sig DBQ Crash 59 1, 148 0.63 .605 .438 No crash 90 0.59 Self-reported driving behaviour Crash 59 1, 148 0.40 .812 .369
No crash 90 0.32 Intended driving behaviour Crash 59 1, 148 0.38 .816 .368 No crash 90 0.30 Subjective norms* Crash 56 1, 137 3.28 1.175 .263 No crash 82 3.39 Attitude Crash 59 1, 148 4.00 .030 .862 No crash 90 3.98 Perceived behavioural control Crash 59 1, 148 5.17 .747 .389 No crash 90 5.04 Beliefs about environmental constraints Crash 59 1, 148 2.45 .000 .986
No crash 90 2.45 * Instrument excluded first 3 items due to large amounts of missing data
6.5.3.4 Knowledge of level crossing road rules
Participants were asked four questions about road rules and facts about level
crossings. Less than a fifth of drivers (14.8%) answered all questions correctly, while
three participants (2.0%) did not answer any questions correctly. Compared to older and
heavy vehicle participants, younger participants scored substantially lower in knowledge.
This is an interesting finding as younger drivers would most likely have been exposed to
some information about level crossing road rules in the licensing process.
6.5.3.5 General driving behaviour
Mean scores for items on the modified driver behaviour questionnaire were very
low (i.e. less than 1) at both time points. The mean score at Time 1 was 0.61
(S.D.=0.338) which remained relatively similar at Time 2 (M=0.77, S.D.=0.329).
6.5.3.6 Self-reported driving behaviour at level crossings
At Time 1, the mean score for this instrument was 0.35 (S.D.=0.530) and increased
to 1.23 (S.D.=1.894) at Time 2. Both these scores are relatively low, indicating that
younger participants believe they drive safely at level crossings.
255
6.5.3.7 Intended driving behaviour at level crossings
Mean scores for driver intention at Time 1 (M=0.33, S.D.=0.565) was very low,
indicating that younger participants report that they intend to drive safely at level
crossings in the future. A correlation analysis was undertaken to explore the relationships
between the dependent variable driving intention at level crossings (Time 1) and the
outcome variables. Table 69 illustrates the significant relationships that were revealed.
As shown in Table 69, there was a weak negative correlation between participants’
gender and driving intention at level crossings (Time 1) [r = -.173, p < .05]. In other
words, being a male participant was associated with a stronger intention to drive unsafely
at level crossings. In addition, there was a significant negative correlation between the
participants’ self-efficacy whilst driving at level crossings and driving intention at level
crossings (Time 1) [r = -.216, p < .01]. This negative relationship indicated that lower
self-efficacy levels were associated with a stronger intention to drive unsafely at level
crossings.
Table 69: Bivariate correlations between dependent and independent variables Dependent Variable
Independent variables Driving Intention at Level Crossings (T1)
Age -.068 Gender -.173 * Educational level -.099 Licence type .069 Rurality .114 Hours per week drive .036 Crashes during past 3 years .074 DBQ .253 ** Attitude .033 Perceived behavioural control -.216 ** Subjective norm .152 Environmental constraints .103 Perceived risk of collision .025 Driving intention - Self-reported driving behaviour .809 **
* p < .05 ** p < .01 *** p < .001
256
To further analyse the prediction of driving intention at level crossings, a
hierarchical multiple regression was used. This 3-step hierarchical regression analysis
was undertaken to assess the contribution of the components of the integrated model (IM)
to the prediction of behavioural intention, along with measures of environment,
skills/ability (compliance with road laws) and perceived risk. Three blocks of variables
were used to predict intention at Time 1: (i) environmental constraints and compliance
with road laws, (ii) attitudes, subjective norms and perceived behavioural control, and
(iii) perceived risk. In this way, it was possible to assess the additional predictive utility
of the attitudes, subjective norms, and perceived behaviour control and the ability of
environmental constraints and compliance with road laws to mediate the influence of
perceived risk.
The overall 3-step regression model was not significant, however perceived
behavioural control (self-efficacy) was found to be a significant predictor of intention to
drive safely at level crossings (p<.01). At Step 1, environmental constraints and
compliance with road laws variables accounted for only 1.6% of the variance in
behavioural intention. The addition of attitudes, subjective norms and perceived
behavioural control (F change = 3.365, p<.05) led to small (8.6%) increases in the
amount of variance explained, however the final addition of perceived risk (F change =
.048, NS) led to even smaller (8.7%) increases in the overall amount of variance
explained by the model.
257
Table 70: Hierarchical regression of constructs on intention at level crossings Variables Mean S.D. B Std
Error β R2 Adj R2 ∆ R2
Step 1 Environmental constraints 2.52 .815 -.011 .061 -.016
Compliance with road laws# 2.99 .471 -.136 .103 -.113
.016 .002 Step 2 Attitude 4.06 .827 .071 .070 .103 Subjective norm 3.35 .788 -.002 .069 -.003 Perceived behavioural control 5.13 .870 -.185 .060 -.284**
.086 .052 Step 3 Perceived risk 2.49 .766 -.014 .064 -.019 .087 .045 .000
* p<.05 ** p<.01 ***p<.001 # Compliance with road laws was used instead of driving ability as this measure is likely to be more accurate in a younger driver sample
6.5.3.8 Attitudes towards driving at level crossings
The first eight items from the 15-item instrument with bipolar scales (i.e. -3 to +3)
directly measured attitudes towards driving at level crossings. As stated previously, each
of the bipolar scales was recoded using unipolar (1-7) scales. Participants were asked to
complete statements by rating pairs of adjectives whilst driving at level crossings (e.g.
bad-good, more confusing-less confusing). Mean scores indicated that at Time 1 their
was generally a positive attitude towards driving at level crossings, however one item
scored less than a mean of 3.0. The item ‘design of crossings (difficult to obey rules to
easy to obey rules)’ had a mean of 2.90 at Time 1. This indicated that younger drivers
believe that it is sometimes difficult to obey the road rules at level crossings due to the
design of crossings. Table 71 illustrates the mean scores for each item.
258
Table 71: Attitudes towards driving at level crossings Time 1
Item n
Mean
Design of crossings (bad) 149 4.21Design of crossings (unsafe) 149 4.13Design of crossings (confusing) 149 3.41Design of crossings (difficult to obey rules) 149 2.90Road rules (bad) 149 4.47Road rules (not strict enough) 149 4.54Road rules (confusing) 149 4.34Road rules (not practical) 149 3.91
1=Negative attitude to 7=Positive attitude
6.5.3.9 Perceived behavioural control whilst driving at level crossings
The second seven items from the 15-item instrument with bipolar scales (i.e. -3 to
+3) directly measured self-efficacy and perceived behaviour control whilst driving at
level crossings. As stated previously, each of the bipolar scales was recoded using
unipolar (1-7) scales. Participants were asked to complete statements by rating pairs of
adjectives whilst driving at level crossings (e.g. not up to me-up to me, out of my control-
under my control). Mean scores at Time 1 indicated that younger participants generally
believe that they have control over their driving at level crossings, however one item
‘other motorists influence my driving’ (more stressful to more relaxing) scored a lower
mean (M=3.96) which indicated that some participants feel that it is sometimes stressful
driving at level crossings because of other motorist’s behaviour. Table 72 illustrates the
mean scores observed at Time 1 for self-efficacy/perceived behavioural control.
259
Table 72: Perceived behavioural control whilst driving at level crossings Time 1
Item n Mean
Obeying the rules (not up to me) 149 4.96Obeying the rules (out of my control) 149 5.30Obeying the rules (dependent on other motorists) 149 5.91Obeying the rules (dependent on time constraints) 149 6.05Other motorists influence my driving (harder to obey rules) 149 5.35Other motorists influence my driving (more confusing) 149 4.11Other motorists influence my driving (more stressful) 149 3.96
1=Negative attitude to 7=Positive attitude
6.5.3.10 Subjective norms of important others
As stated previously, the first three items in this instrument comprised large
amounts of missing data, possibly due to online electronic data saving problems.
Therefore, the first three items of this instrument were not included in mean scores.
Table 73 below illustrates the mean scores for Time 1. The lowest mean score ‘generally
safe to roll slowly through a crossing instead of stopping’ (M=1.90) suggests that
younger drivers are prepared to take a risk and not stop at crossings.
Table 73: Mean scores for subjective norms of others Time 1
Item n Mean
Your family generally obeys the rules at rail crossings - -Your friends generally obey the rules at rail crossings - -Other motorists generally obey the rules at rail crossings - -Your family generally think it important to obey the rules at rail crossings 143 4.30Your friends generally think it important to obey the rules at rail crossings 142 5.30Other motorists generally think it important to obey the rules at rail crossings 142 4.68
It is generally safe to disobey the rules at rail crossings 142 4.54It is generally possible to judge a train’s speed 142 2.86It is generally safe to cross if you can’t see a train, even if the lights are flashing 141 2.82
It is generally safe to roll slowly through a crossing instead of stopping 142 1.90Trains generally run to a regular timetable 142 2.01Penalties need to be tougher for violating road rules at rail crossings 141 2.51The main deterrent for breaking the rules at rail crossings is fear of getting caught 141 3.41
Generally it is more important to use common sense at rail crossings than strictly follow the road rules 141 2.65
1=Strongly agree to 6=Strongly disagree
260
6.5.3.11 Perceived risk of a level crossing collision
Younger participants were asked about their perception of how likely it is that they
would be involved in a collision at a level crossing whilst driving. The younger sample
mean score (M=2.42) at Time 1 were relatively higher than both older drivers (M=0.19)
and truck drivers (M=0.33). There appeared to be a perception among younger drivers in
this sample that they are at risk of being involved in a level crossing collision.
6.5.3.12 Environmental constraints whilst driving at level crossings
At Time 1, the items ‘design of the road makes it difficult to see if a train is
approaching’ (M=3.12) and ‘difficult to hear an approaching train when the windows are
up’ (M=3.02) scored the highest means for younger drivers. The lowest mean observed
at Time 1 was ‘my car stalled on the tracks at a rail crossing’. Table 74 presents Time 1
mean scores for environmental constraints.
Table 74: Beliefs of environmental constraints whilst driving at level crossings Time 1
Item n Mean
Blinding sun makes it difficult to see if the red flashing lights are activated 149 2.35The design of the road makes it difficult to see if a train is approaching or at the rail crossing 149 3.12
Warning systems on the road approaching the crossing are not adequate to inform drivers there is a rail crossing ahead 149 2.71
Road surfaces are often poor and it is difficult to stop 149 2.35Boom gates and/or flashing lights are often faulty 149 1.99Other drivers do stupid things that put you in a dangerous situation 149 2.99Intersections ahead of a rail crossing often cause your car to overhang the tracks 149 2.29
Difficult to hear an approaching train when the windows are up 149 3.02My car stalled on the tracks at a rail crossing 149 1.19
1=Never, 2=Almost never, 3=Sometimes, 4=Almost always, 5=Always
6.5.3.13 Familiarity with different protection systems
Table 75 illustrates participant reported exposure to driving at level crossings. The
majority of participants reported that they never drive through boom gates (53.0%), never
drive through crossings with flashing lights (58.4%) and never drive through passive
261
crossings without lights or boom gates (52.3%). As such, it is evident that more than half
of the participants were more likely to be unfamiliar with driving at level crossings.
Table 75: Exposure to level crossing driving Boom gates Flashing Lights only No lights or boom gate
Type n % n % n % Never 79 53.0 87 58.4 78 52.3Once a year 11 7.4 21 14.1 22 14.8Twice a year 17 11.4 15 10.1 9 6.0Monthly 15 10.1 14 9.4 8 5.4Weekly 17 11.4 12 8.1 10 6.7Daily 10 6.7 0 0 22 14.8
Like the other two road user groups, the exposure instrument was then recoded to
compare those participants that either drive through each type of crossing regularly or
not. Those participants that stated they never, yearly or twice a year drove over either
active or passive crossings were combined to form a group of being unfamiliar with that
type of crossing whilst driving. Surprisingly, only 4% of participants that lived in
metropolitan areas were familiar with crossings that had only flashing red lights, while
similar percentages for familiarity with passive crossings (stop or give-way signs) were
reported for metropolitan and rural/remote drivers.
Table 76: Familiarity and area classification at Time 1 Familiar Unfamiliar
Protection System n % n % Boom gates Metropolitan 27 18.1 65 43.6 Rural and Remote 15 10.1 42 28.2 Flashing red lights Metropolitan 6 4.0 86 57.7 Rural and Remote 20 13.4 37 24.8 Give-way or Stop sign only Metropolitan 19 12.8 73 49.0 Rural and Remote 21 14.1 36 24.2
262
A series of one-way analysis of variances were calculated to examine differences
between familiarity of younger drivers with level crossings (boom gates, flashing lights
or passive) in relation to each of the driving instruments. Drivers that were considered
‘familiar’ with level crossings were those that reported driving over a type of crossing
daily, weekly or monthly. Results are presented in the tables below.
Table 77: Familiarity with driving at level crossings (boom gates) Instrument Familiarity n df Mean F Sig DBQ Familiar 42 1, 148 0.67 1.869 .174 Unfamiliar 107 0.58 Self-reported driving behaviour Familiar 42 1, 148 0.55 8.346 .004* Unfamiliar 107 0.27 Intended driving behaviour Familiar 42 1, 148 0.46 3.241 .074 Unfamiliar 107 0.28 Beliefs about others’ driving* Familiar 39 1, 137 3.25 1.052 .410 Unfamiliar 99 3.39 Attitudes towards driving at level crossings Familiar 42 1, 148 4.09 .841 .361
Unfamiliar 107 3.95 Self-efficacy whilst driving at level crossings Familiar 42 1, 148 4.99 .746 .389
Unfamiliar 107 5.13 Beliefs about environmental constraints Familiar 42 1, 148 2.54 .786 .377
Unfamiliar 107 2.41 * Instrument excluded first 3 items due to large amounts of missing data
The instrument on self-reported driving behaviour at level crossings was the only
significant factor for familiarity with boom gated crossings [F (1, 148) = 8.346, p <0.05].
This indicated that younger drivers that were familiar with driving at crossings with boom
gates recorded higher levels of unsafe driving at level crossings generally than unfamiliar
drivers.
263
Table 78: Familiarity with driving at level crossings (flashing lights) Instrument Familiarity n df Mean F Sig DBQ Familiar 26 1, 148 0.73 4.432 .037* Unfamiliar 123 0.58 Self-reported driving behaviour Familiar 26 1, 148 0.53 3.498 .063
Unfamiliar 123 0.31 Intended driving behaviour Familiar 26 1, 148 0.53 3.886 .051 Unfamiliar 123 0.29 Beliefs about others’ driving* Familiar 25 1, 137 3.37 .977 .516 Unfamiliar 113 3.34 Attitudes towards driving at level crossings Familiar 26 1, 148 4.33 5.142 .025*
Unfamiliar 123 3.92 Self-efficacy whilst driving at level crossings Familiar 26 1, 148 5.23 .705 .402
Unfamiliar 123 5.06 Beliefs about environmental constraints Familiar 26 1, 148 2.58 .743 .390
Unfamiliar 123 2.42 * Instrument excluded first 3 items due to large amounts of missing data
A significant effect for two instruments: driver behaviour questionnaire F (1, 148)
= 4.432, p <0.05 and attitudes towards driving at level crossings [F (1, 148) = 5.142, p
<0.05] were revealed. This indicated that younger drivers that were familiar with driving
at crossings with only flashing lights recorded a higher level of unsafe driving generally
but had a more positive attitude towards driving at level crossings.
264
Table 79: Familiarity with driving at level crossings (passive signs only) Instrument Familiarity n df Mean F Sig DBQ Familiar 40 1, 148 0.71 5.660 .019* Unfamiliar 109 0.57 Self-reported driving behaviour Familiar 40 1, 148 0.49 4.042 .046 Unfamiliar 109 0.30 Intended driving behaviour Familiar 40 1, 148 0.46 2.829 .095 Unfamiliar 109 0.29 Beliefs about others’ driving* Familiar 38 1, 137 3.31 1.037 .431 Unfamiliar 100 3.37 Attitudes towards level crossings Familiar 40 1, 148 4.14 1.779 .184
Unfamiliar 109 3.93 Self-efficacy whilst driving at level crossings Familiar 40 1, 148 5.14 .182 .671
Unfamiliar 109 5.07 Beliefs about environmental constraints Familiar 40 1, 148 2.50 .215 .643
Unfamiliar 109 2.43 * Instrument excluded first 3 items due to large amounts of missing data
A significant result was found for the instrument ‘driver behaviour questionnaire’
[F (1, 148) = 5.660, p <0.05] and familiarity with passive crossings (stop or give-way
sign only). This indicated that younger participants that are familiar with driving at
passive crossings are more likely to drive unsafely on the road generally than participants
that reported being unfamiliar with passive crossings.
6.5.3.14 Intervention message recall
Recall of what message younger drivers received was better than the two other road
user groups. Fifty-seven (64.7%) recalled either receiving the message about wearing
seatbelts at all times when driving or safe driving at level crossings. Only three (3.4%)
participants recalled any of the slogans from either of the messages, while eighteen
participants (20.5%) recalled information about either of the messages. Only four (4.5%)
participants reported that by listening to the road safety messages their driving behaviour
has been affected.
265
6.5.3.14 Effect of intervention exposure
A comparison of means between the intervention and control groups was
undertaken at both Time 1 and Time 2. Table 80 below illustrates the mean scores for all
outcome variables at Time 1 and Time 2 for both the intervention and control groups for
the younger driver sample. Mean scores for the driver behaviour questionnaire increased
for both the intervention and control groups from Time 1 to Time 2, indicating that those
participants that remained in the study at Time 2 were more prepared to take risks on the
road than the Time 1 younger driver sample. Additionally, mean scores for attitudes
towards driving at level crossings also increased for both intervention and control groups,
however this indicated that the Time 2 sample had more positive attitudes towards
driving at level crossings. Also of importance is the considerable decrease in the mean
scores for perceived risk of a collision at a level crossing, which indicates that the sample
of participants that remained at Time 2 were less likely to believe that they are at risk of a
level crossing collision.
Table 80: Comparison of experimental group instrument mean scores Intervention Group Control Group
Outcome Variables
T1 Mean
T2 Mean
T1 Mean
T2 Mean
Driver behaviour questionnaire .64 .75 .65 .80Self-reported driving behaviour .35 1.26 .45 1.17Intended driving behaviour .32 .30 .40 .36Attitudes towards driving at level crossings 4.09 4.41 4.11 4.36
Self-efficacy of driving at level crossings 5.12 4.55 4.96 4.73
Normative beliefs of others driving at level crossings* 3.25 3.21 3.51 3.00
Beliefs of environmental constraints whilst driving at level crossings 2.44 2.62 2.74 2.77
Perceived risk of a collision at a level crossing 2.53 1.29 2.40 1.30
* First 3 items removed from both Time 1 and Time 2 mean scores due to missing data in Time 1
To ascertain the effect of the intervention for the younger driver sample, a split-plot
in time analysis (repeated measures) was undertaken for each of the outcome variables.
Separate analyses were undertaken rather than multivariate ANOVAs since the dependent
266
variables were predicted to have different outcomes and the outcomes of interest were the
individual variables rather than the set of variables.
Results of the series of split-plot analyses reveal that normative beliefs (F = 6.196=
1, 81P< 0.05) was the only outcome variable to have significantly changed over time
between groups. The associated Partial Eta squared was .071 indicating a small to
medium effect size. A number of outcome variables (see Table 81) were observed to have
a significant time effect (within-subject test), however between groups tests indicates that
the variable groups were not significant. Results from the repeated measure analyses are
presented in the table below.
267
Table 81: Repeated measure analysis of variance for the outcome variables Instrument df Wilks’ A
Effect sizes(Partial Eta
squared)F Sig
DBQ
Time 1, 86 .793 .207 22.434 .000***
Group x Time 1, 86 .993 .007 .606 .438
Self-reported driving behaviour
Time 1, 86 .871 .129 12.774 .001**
Group x Time 1, 86 .998 .002 .169 .682
Intended driving behaviour
Time 1, 86 .996 .004 .358 .551
Group x Time 1, 86 .999 .001 .058 .811
Attitudes
Time 1, 86 .940 .060 5.520 .021*
Group x Time 1, 86 .999 .001 .088 .768
Self-efficacy
Time 1, 86 .831 .169 17.436 .000***
Group x Time 1, 86 .964 .036 3.252 .075
Normative beliefs
Time 1, 81 .908 .092 8.180 .005**
Group x Time 1, 81 .929 .071 6.196 .015*
Environmental constraints
Time 1, 86 .985 .015 1.290 .259
Group x Time 1, 86 .993 .007 .565 .454
Perceived risk of a collision Time 1, 86 .688 .312 38.952 .000***
Group x Time 1, 86 .998 .002 .142 .707
* p<.05 ** p<.01 ***p<.001 Note: to interpret effect sizes: .02=small magnitude; .15=medium magnitude; .35=large magnitude (Cohen, 1988).
268
6.6 STUDY LIMITATIONS
6.6.1 Widespread limitations
Like all research, this final study is not without its limitations. There are
widespread limitations in the data collected from each of the three road user groups,
as well as specific limitations of the data collection methodologies within these
groups.
Firstly, one of the widespread limitations of this final study is that data for all
three road user groups was drawn exclusively from only one jurisdiction in Australia.
Every effort was made to ensure that there was a diverse sample of participants in
both the older and younger driver groups, with an equal number of recruitment letters
being sent to rural/remote and metropolitan RACQ members. Additionally,
randomisation to either the intervention or control group was conducted for
rural/remote and metropolitan participants. With financial constraints limiting the
scope of data collection extending outside of Queensland, some caution needs to be
exercised when generalising these results to other jurisdictions in Australia or abroad.
However, anecdotal evidence from train drivers throughout Australia suggests that
there is very little difference in the behaviours and road user groups most at risk of a
collision with a train.
Secondly, this investigation suffers from the typical perceived limitation
associated with self-report data. The main criticism of self-report data is that it is
affected by social desirability bias. It goes without saying in road safety research that
social desirability will have some impact upon the findings obtained. It has been
suggested that ‘deviant’ individuals would minimise the number of crashes they were
involved in or other illegal driving activities (e.g. drink driving) (Hatakka, 1997).
However, it is important to note that any resulting associations found would be under-
estimates of real associations, rather than over-estimates (Hatakka, 1997, Lawton,
1997a, West, 1995). For the current investigation, the use of self-report would have
potentially been a problem if actual (observed) driving behaviour was monitored.
However, since actual driving behaviour was not monitored, the degree of
correspondence between self-report and observed driving behaviour is insignificant.
Thirdly, the use of incentives was considered to be suitable for each of the three
road user groups, however may not have been great enough for completion of both
269
questionnaires. Two movie tickets were provided to each of the participants if they
completed both questionnaires. This may have been a great enough incentive to
originally agree to participate, but over time more incentives may have been needed to
ensure low attrition rates. However, limited financial resources did not allow four
movie tickets to be purchases for all participants (at a total cost of approximately $44
per participant).
Finally, as this intervention was provided as a ‘one-off’ intervention, it appears
that the recall was extremely poor. The intervention would have been enhanced if
participants were exposed to the intervention on several occasions over a period of a
few weeks, rather than just once. However, again due to the limited budget in which
the intervention could be rolled out, a ‘one-off’ brief intervention was the most
achievable method for delivery. A larger budget, such as that spent by governments
in developing road safety awareness campaigns, would have possibly achieved greater
recall of this intervention as participants could have been exposed to a range of
delivery modes (i.e. radio, television, billboards) over a longer period of time.
6.6.2 Heavy vehicle driver sample limitations
Data collection from the trucking industry was fraught with difficulty. The
greatest challenge was the poor involvement of companies in the study. Although the
project sought the assistance of the Queensland Trucking Association and actively
approached a multitude of trucking companies, only five agreed to participate. This
was an extremely disappointing response rate, particularly with the months of effort in
approaching companies and offering incentives to participants. To the author, it
appears that the reluctance by many companies to participate in this road safety
research may have stemmed from either the belief that level crossing safety is not
high on their safety agenda or a distrust of disclosing driving behaviours of employees
in fear of scrutiny of the company’s safety compliance. The lack of involvement
from the trucking industry therefore resulted in a relatively small sample size that may
limit the validity and generalisability of the findings, as well as the range of statistical
analyses that could be undertaken. It could also be argued that those companies that
were willing to participate, had better safety records that those companies that refused
to participate. As such, this study may have only captured the companies with
270
comprehensive OHS management systems and a focus on ensuring a culture of safety
within their workplace.
Additionally, this small sample size prevented randmonisation of companies to
either the intervention or control group from occurring. This was a major limitation
for this road user group and further techniques and incentives for company participant
may need to be developed in order to increase the industry’s participation rates over
longer periods of time.
The administration of the questionnaire was another area that limited the
response rate from truck drivers. Although the first questionnaire (Time 1) was
completed face-to-face by drivers during morning tea breaks at their depot, the Time 2
questionnaire was posted to the drivers. This was deemed the most cost-effective data
collection tool for this subsequent questionnaire, however it appears that truck drivers
are less willing to complete and return questionnaires outside of their workplace. As
there was a significant attrition between Time 1 and Time 2, it can therefore be
assumed that it would have been more effective for participants to have face-to-face
involvement with completion of both questionnaires. Mean safety scores on many of
the items in the questionnaire are higher at Time 2 than they are at Time 1 which
would suggest that drivers that completed both questionnaires were likely to drive
more safely than those that only completed the first questionnaire. Hence, it could be
said that participants that were willing to complete both questionnaires had an interest
in road safety as well as level crossing safety. Although the author sent reminder
letters to those participants that hadn’t returned the questionnaire within 3 weeks of
being posted, focus group research at the truck depots may have assisted with further
data collection for this group. However, the lack of financial resources did not allow
this to occur.
Fourthly, although a modified version of the DBQ was developed for this
sample of truck drivers, the purpose of the original DBQ was to measure driving
behaviours specifically amongst car drivers, not truck drivers. According to Sullman
et al (2002), the vehicle dynamics of a truck differ greatly to that of cars and therefore
this may influence the types of driving behaviour that truck drivers engage in. As
such, this must be kept in mind when interpreting the findings from the DBQ with the
heavy vehicle sample. Finally, although self-report data generally has limitations
(such as social desirability bias), a study of New Zealand truck drivers comparing
self-report speeds with mean speeds observed by the Land Transport Safety Authority
271
(LTSA), found that truck drivers accurately reported the speeds at which they traveled
on the road (Walton, 1999). Such studies therefore provide support for the validity of
self-report measures of driving behaviour for heavy vehicle drivers.
6.6.3 Older driver sample limitations
One of the most obvious limitations with the older driver sample is that there
were several couples (i.e. husband and wives) that participated in the research. It
remains unknown if these couples completed the questionnaires entirely on their own,
or whether they talked about the questionnaires with each other during completion.
Although all participants were informed that they should complete the questionnaire
by themselves, no guarantees can be made about this as they were postal
questionnaires rather than one-on-one interviews. Additionally, each participant was
randomly allocated to receive either the intervention or control road safety message
(via telephone). Once again, although all participants were asked not to discuss what
message they received with their spouse, it is unknown if any couple talked about
what road safety message they heard when contacted by a research assistant.
Consequently, there may a small amount of contamination with questionnaires that
were completed by couples.
6.6.4 Younger driver sample limitations
There were two leading limitations with the younger driver sample. Firstly,
there was some undiscovered online survey saving problems (with one particular
instrument) that were not detected until both databases were received from the
external online survey company. As such, it was too late to contact participants to re-
enter their data for this instrument. It appeared that one particular page did not save
correctly for some participants, and consequently this instrument was not able to be
included in pre and post-test comparisons for this younger driver sample. The
discovery of this problem was very disappointing, as the author was particularly
interested in this instrument for the younger driver sample. The second limitation
for this road user group was that of participants completing more than one survey (at
both time points) to try to gain more movie tickets. As the data was cleaned and
checked rigoursly for any duplicate information (such as personal information like
272
email addresses), approximately half a dozen duplicate entries were discovered. This
resulted in the sample being slightly smaller than originally thought to be when the
databases were received. This was certainly an interesting find by the author,
however was not surprising given that movie tickets were deemed by many younger
drivers (during focus group discussions) as being a great incentive for participation.
6.7 SUMMARY
With the assistance of Fishbein’s Integrated Model of Behaviour Change (IM),
an exploratory investigation of three different road user groups was achieved. Two
aims and eight research questions were addressed in the final study of the research
program. This summary outlines the findings relevant to each of the aims and
research questions.
The first aim was to examine what changes in driving intention and self-
reported driving behaviour are influenced by exposure to the intervention message.
Research Question 1: Does exposure to the intervention message produce safer
driving intention and self-reported driving behaviour at level crossings?
There were no significant changes to either driving intention or self-reported
driving behaviour in any of the three road user groups. However, given that the
intervention was a ‘one-off’ intervention due to limited financial resources for
implementing a large scale education program such as those developed by transport
authorities, this result was expected.
The second aim was to examine the personal, social and environmental factors
that influence driving intention to perform behaviours related to unsafe driving at
level crossings.
Research Question 2: Is there a relationship between gender and unsafe driving
intention among older and younger drivers?
A relationship between gender and unsafe driving intention was only found for
the younger driver sample. Young male participants were significantly more likely to
report both current and future driving at level crossings to be less safe than their
female counterparts. This is not surprising given the amount of literature supporting
273
research that young males are more likely to take risks whilst driving generally
(Laapotti and Keskinen, 1998, Arnett, 2002) or be less ‘safety-orientated’ than their
female counterparts (Meadows and Stradling, 1999). However, this finding has
implications for targeting younger drivers as it appears that educational programs may
need to direct efforts largely towards young male drivers as opposed to their female
counterparts. The challenge remains for program designers to implement
interventions that have the prospect to produce lasting behaviour change in young
males.
Research Question 3: Is there a relationship between distance traveled per week
and unsafe driving intention among heavy vehicle drivers?
A relationship was found between distance traveled per week and unsafe driving
intention in the heavy vehicle driver sample. A strong association was found between
the number of kilometres driven per week in a heavy vehicle and driving intention at
level crossings. Although it is unknown if the participants that reported longer
distances traveled were abiding by fatigue management policies and practices by
exceeding prescribed driving hours, a number of safety issues are known to be
associated with increased driving exposure. “A direct consequence of the long hours
worked by drivers is to resort to stimulant drugs. Drivers use drugs not for pleasure
but to combat fatigue and stay at the wheel longer” (Quinlan, 2001, p4). Research
conducted by Williamson et al (2003b) found that there is a strong association
between tight schedules, delivery time bonus/penalties and performance-based
payment systems (e.g. kilometre-based rates) and both chronic injury and the
propensity of drivers to engage in dangerous driving practices (such as speeding and
excessive hours). Given that speeding heavy vehicles are significantly over-
represented in crashes, the finding from the current research requires further
investigation.
Research Question 4: Is there a relationship between familiarity and unsafe
driving intention?
Familiarity with driving at level crossings was found to be associated with
unsafe future driving intention within all three samples. Some literature has suggested
that familiarity may encourage motorists to take greater risks (Abraham, 1998).
According to Pickett (1996) drivers may “also base their decision to cross on their
274
previous experience of either the same crossing (familiarity) or of other crossings
(association). The concept of mental set says that when people are exposed to the
same phenomenon repeatedly, they come to expect it” (p5). Pickett (1996) also
suggests another situation, with drivers familiar with one crossing, transferring their
experience to a new crossing and not being as vigilant in the new situation. However,
Wigglesworth (1979) proposes that determining an accurate picture of familiarity with
a particular level crossing is difficult. In Wigglesworth’s study of 85 consecutive
fatalities in Victoria (between 1973 and 1977), he found that 73 of the 85 motorists
(86%) that were fatally injured were considered to be familiar with the crossing at
which their collision occurred (Wigglesworth, 1979). These motorists lived within
one mile of the level crossing. The finding that familiarity is associated with higher
levels of risk taking has significant implications for all three road user groups, but
particularly the heavy vehicle industry given the potential for a major catastrophic
event to occur if a truck collides with a train.
Research Question 5: Is there a relationship between attitudes and unsafe
driving intention?
No relationship between attitudes and unsafe driving intention was found for
any of the driver samples. Attitudes are defined as a person’s overall evaluation of a
particular behaviour leading to either positive or negative outcomes (Lajunen, 2004a).
Given that some research has suggested that attitudes have a greater influence on
driving intentions in a ‘safe’ driving environment while normative influences were
more influential in a ‘risky’ driving environment (Trafimow and Fishbein, 1994), this
finding is not surprising. The current research findings support Trafimow and
Fishbein’s (1994) investigation suggesting that normative factors appear to be more
influential in intentions to drive unsafely at level crossings by all three road user
groups than attitudinal factors.
Research Question 6: Is there a relationship between self-efficacy and unsafe
driving intention?
A relationship between self-efficacy and unsafe driving intention was found for
both the older and younger driver samples. Self-efficacy has been defined as an
individual’s capacity to organise, control and execute certain behaviours to attain
specific performances (Bandura, 1977a). However, since self-efficacy is regarded as
275
a coderterminant of behaviour (Ajzen, 1991), it relies on the accuracy of an
individual’s perception of their control over a driving situation (Sheeran, 2003). Self-
efficacy was shown to be central for both the younger and older driver samples. Older
drivers reported lower levels of self-efficacy of driving at level crossings as the
majority believe that other motorists influence their driving and their ability to obey
the road rules. Such findings however contradict previous research into speeding.
Elliot et al (Elliott, 2003b) in their research found that older drivers had greater
perceived behavioural control (perceived self-efficacy) then younger drivers when it
came to complying with speed limits. Self-efficacy was also found to be associated
with the driving intention of younger drivers. The current research produced similar
results to previous research with younger drivers. Elliot et al (Elliott, 2003b) in their
research about driver’s compliance with speed limits found that younger drivers (and
particularly males) had lower levels of self-efficacy than older drivers. Some research
has indicated that individuals with higher levels of self-efficacy report significantly
greater rates of success in changing behaviour (Wells-Parker et al., 2000). However,
one of the major assumptions of Fishbein’s Integrated Model (IM) is that self-efficacy
has primarily an indirect effect on behaviour (via their effects on intention) (Fishbein
et al., 2003). Thus, if young drivers have already formed strong intentions to perform
unsafe driving at level crossings (but are not doing so), then little will be
accomplished by developing interventions designed to increase self-efficacy (Fishbein
et al., 2003). This has important implications for designing educational programs for
young drivers.
Research Question 7: Is there a relationship between subjective norms and
unsafe driving intention?
A relationship was found between subjective norms and unsafe driving intention
for both the older and younger driver samples. Subjective norm consists of a person’s
beliefs about whether significant others (such as friends, family or colleagues) think
that they should engage in a particular behaviour (Conner, 1996). Older driver
participants hold the belief that other motorists do not obey the rules at level crossings
as often as their important others (i.e. family and friends). This finding also supports
the formative research undertaken with older drivers in that they hold the belief that
their age group drives more safely at level crossings than other road user groups.
Although older drivers are generally not characterised by risk-taking behaviours, their
276
driving errors reflect functional, cognitive and perceptual impairment (Eby, 1998b).
Although the current research is not able to confirm causal relationships per se, it has
highlighted that prevalence of functional impairments in older drivers aged 60 plus
years. Further research should be directed at examining any associations between
cognitive and perceptual impairment for driving at level crossings for this age group.
For the younger driver sample, there was a small to medium effect size found
for normative beliefs. The observed shifts in normative beliefs for this age group (17-
24 years) may possibly be due to the perceived risk of the driving environment at
level crossings. Some research has suggested that attitudes have a greater influence
on driving intention in a ‘safe’ driving environment, while normative influences are
more influential in a ‘risky’ driving environment (Trafimow and Fishbein, 1994). Of
the three road user groups, younger drivers hold a greater perception of risk whilst
driving at level crossings. Therefore, it could be argued that the level crossing
environment is perceived a ‘risky’ driving environment by younger drivers, which
would explain why normative beliefs are significantly changed over time compared to
other outcome variables.
Research Question 8: Is there a relationship between beliefs about
environmental constraints and driving intention?
Although the heavy vehicle group reported greater frustration with
environmental factors such as the design of the approaching road and flashing lights
being difficult to see with the sun, there was no association with driving intention in
any of the three samples.
The subsequent chapter will combine the findings of all three studies and
endeavor to amalgamate the implications of the research in pursuit of attaining an
improved understanding of motorist behaviour at level crossings. Such pursuits will
hopefully create a more comprehensive approach to the future development of
effective educational countermeasures. This chapter also considers the contribution of
the research findings to the management of risk at level crossings in Australia.
277
CHAPTER SEVEN: DISCUSSION
7.1 Introduction …………………………………………………………… 278
7.2 Overview of the studies, methodology and key findings ……………... 280
7.3 Implications for level crossing safety …………………………………. 296
7.4 Strengths and weaknesses of the research …………………………….. 300
7.5 Suggestions for future research ……………………………………….. 302
7.6 Recommendations to industry ………………………………………… 302
278
7.1 INTRODUCTION
Level crossing collisions in Australia are a major cause of concern for both rail
and road authorities. When a catastrophic collision occurs (such as a heavy vehicle-
train collision resulting in a derailment with numerous fatalities), it receives great
media attention, and the actions of rail/road infrastructure providers and the behaviour
of motorists comes under close scrutiny. It must be noted though that as long as level
crossings present inherent dangers to motorists (and as long as they exist), rail
authorities will remain at risk of being found liable (either in whole or in part) for
fatal collisions (Stephen, 2002). With more than 9400 level crossings throughout
Australia and limited funds for upgrading crossings, it is an arduous task for rail
authorities to provide engineering systems that will completely protect all motorists
from involvement in a collision. Historically, research in the level crossing arena has
focused on engineering solutions to protect both rail and road users, however more
recently, the rail industry has sought new approaches to improving risk management
initiatives. One such approach is the examination of motorist behaviour at level
crossings. To date, little rigorous research has been conducted that has either
explored the determinants of behaviour and associated key constructs or measured
changes in behavioural constructs or statistically significant interactions between such
constructs. Additionally, no research has been guided by psychological theory in
exploring motorist behaviour at level crossings. As a consequence, little is known
about the driving intentions of motorists and their beliefs of factors influencing their
driving in these complex environments. Without such research, it is difficult for
transport authorities to justify expenditure on campaigns targeting behaviour.
The ultimate goal to improve level crossing safety would be to have a
combination of engineering, education and enforcement countermeasures for all level
crossings. However, the small number of fatalities in comparison to the national road
toll limits this. To put level crossings into perspective, such collisions are a
significant problem for the rail industry, yet they represent less than 1% of the
national road toll and are rare events. However, they remain tragic events that cause
tremendous personal, social, financial and infrastructural anguish. Since vehicle-train
collisions are rare events, it is difficult to obtain valid information about preceding
behaviour and the frequency of near-misses. However, unintended road user error has
been blamed by transport authorities as playing a significant role in such collisions.
279
Anecdotal evidence also suggests that risk taking at level crossings makes up a large
proportion of near-miss incidents. It must be noted that the act of driving is taken for
granted by most motorists, however it is a complex task that is dependent on a variety
of cognitive and psychomotor performance abilities to be intact, such as alertness,
attention, multitasking, memory, coordination, and visual spatial perception (Moller,
2004). Hence, from a systematic point of view, level crossing collisions are far more
complex than the immediate causal factor being the non-observance of the road rules.
This background provided the foundation for developing and guiding the three studies
conducted in this program of research.
The primary aim of the research program was to investigate the present context
of motorist behaviour at level crossings in relation to key variables of attitudes,
norms, self-efficacy (perceived behavioural control), perceived risk, environmental
constraints and the skills/abilities of drivers. Three studies were conducted to
investigate high risk and vulnerable road user groups, the relationship between
behavioural constructs, and changes in driving intention after exposure to an
intervention. Both qualitative and quantitative data collection methods were utilised to
provide methodological triangulation and reliability of the data. Prior to this research
program, there have been no documented attempts to investigate the present context
of motorist behaviour at level crossings or the usefulness of targeted educational
interventions for specific road user groups. Given the legal obligations of rail and road
authorities, as well as the public’s scrutiny of protection systems at level crossings, a
greater understanding of motorist behaviour and the context of safe driving at level
crossings remains an important priority for rail/road authorities, legislators,
investigators and researchers alike.
In this chapter, the foremost empirical findings from this research program are
examined and interpreted in relation to the exploration of Fishbein’s integrated model
(IM), as well as practical implications for the future development of government
‘campaigns’ to improve motorist vigilance and promote abidance of road rules at level
crossings. Furthermore, a summary outlining the findings relevant to each of the aims
and research questions is presented.
280
7.2 OVERVIEW OF THE STUDIES, METHODOLOGY AND KEY FINDINGS
A variety of data collection methods were used in this research program as
much of what is currently known about level crossing crashes is derived from
coroner’s findings and crash statistics. Since there are relatively few fatalities per
year (in relation to other fatal road crashes) it is difficult to determine patterns of
contributing factors or high risk road user groups. Therefore, a variety of data
collection measures were utilised to ensure ‘methodological triangulation’ (i.e. the use
of more than one method to investigate a phenomenon).
Study One was designed to extend the current knowledge of fatal collisions at
level crossings by undertaking a more thorough examination of contributing factors to
level crossing crashes and the road user groups at risk. This study applied the
principles of methodological triangulation by using both quantitative and qualitative
research methods. Quantitative data was collected using a modified Delphi technique,
whilst qualitative data was collected using focus group discussions. With the
discipline of road safety research requiring methodological strategies that will
enhance efforts to conceptualise the multi-faceted nature of motorist behaviour at
level crossings, this application provided the robustness required.
Study Two involved formative research as part of the planning, development
and delivery of behavioural interventions for each of the three road user groups
identified in Study One. This study also used both qualitative and quantitative data
collection methods to provide methodological triangulation and ensure reliability of
the data. The overall objective of the qualitative data collection was to obtain rich data
using a qualitative mode of inquiry, based on the key variables of attitudes, norms,
self-efficacy (perceived behavioural control), perceived risk, environmental
constraints and the skills/abilities of drivers. The overall objective of the quantitative
data collection was to prioritise the issues identified in order to direct and allocate
project resources for intervention planning, development and delivery. This combined
recruitment strategy was adopted as it was an appropriate and practical data collection
strategy within the qualitative and exploration methodology. Information obtained
from each of the groups was critical in assisting, guiding, and identifying priority
areas for message and material development. The use of focus groups and one-on-one
interviews provided insights into why drivers think or do what they do at level
crossings.
281
Study Three involved three parts. The aim of Part One of this study was to
develop targeted interventions specific to each of the three road user groups using
Fishbein’s integrated model. The development of interventions was originally seen as
being outside of the scope of this project, however it became intertwined in
questionnaire development and thus deemed to be within the realms of the current
mode of inquiry. The interventions were designed in the format of a pilot radio road
safety advertisement, as this medium was found to be one of the most acceptable to
each of the road user groups as identified in the formative research undertaken in
Study Two. The interventions were used as a ‘one-off’ awareness raising intervention
for each road user group. Part Two involved the investigation of the present context
of unsafe driving behaviour at level crossings. This second part involved the
examination of the present context of motorist behaviour at level crossings using key
constructs from Fishbein’s Integrated Model of Behaviour Change (IM). This study
did not assess the capacity of Fishbein’s model, but rather used it to assist in
exploration of the issues surrounding unsafe driving at level crossings. This issue is
also an important one to examine as no research has ever been undertaken in either
Australia or overseas that has explored the determinants of behaviour and associated
key constructs. Part Three involved trialing a pilot road safety radio advertisement
using an intervention and control methodology. This part investigated the changes in
pre and post-test constructs including intentions, self-reported behaviour, attitudes,
norms, self-efficacy/perceived behaviour control, perceived risks, environment
constraints and skills/ability. This issue is an important one to address as there is no
documented literature that has examined educational interventions for level crossing
safety by explaining intention using a psychosocial theoretical model.
7.2.1 High risk and vulnerable road users at level crossings
Study One provided evidence that there are three main road user groups at risk
of being involved in a level crossing collision. Data collected via a modified Delphi
technique (panel of experts) established that there are four road user groups at ‘high
risk’ of a level crossing collision: heavy vehicles, older drivers, younger drivers, and
rural drivers. Additionally, there were four key areas that experts believed were major
motorist factors that contribute to collisions at level crossings. These included:
• Intentional behaviour – disobeying road rules and trying to beat the train;
282
• Unintentional behaviour – inattention, distraction, complacency;
• Training – drivers are not trained to negotiate level crossings; and
• Understanding – consequence and severity of collisions with trains.
For each of the road user groups identified by the modified Delphi technique
(panel of experts), a number of items that may contribute to increased risk were
presented for ranking. For the older driver group, there was a high level of agreement
that errors in judgment (e.g. misjudging time needed to cross safely) were ‘very
important/important’ risk factors. For the younger driver group, nearly all of the
experts ranked trying to beat the train across the crossing as ‘very
important/important’. Factors for the heavy vehicle group that were ranked as ‘very
important/important’ included: trying to beat the train across the crossing and length
of vehicle causing overhang on the crossing. Factors particular to rural road users that
were ranked as ‘very important/important’ included: low expectation of a train;
complacency due to familiarity and not scanning for a train at give way signed
crossings.
Data collected from focus group discussions with train drivers in both
metropolitan and regional areas revealed that there are strong differences in
experiences of motorist behaviour at level crossings. The metropolitan train drivers
generally experienced motorist behaviour at active crossings with flashing lights and
boom gates while the regional train drivers experienced behaviours at active crossings
with boom gates, crossings with lights only and passive crossings with stationary
signs. In the metropolitan train driver group, experiences of motorist behaviour at
level crossings included: motorists driving around boom gates, getting stuck under
boom gates, queuing over congested crossings and driving through the crossing after
the red lights commence flashing. The behaviour of motorists driving around boom
gates was noted to occur quite regularly. The majority of metropolitan train drivers
reported that it was a common occurrence for motorists to drive through a crossing
when the lights are flashing both before and after the booms were activated and some
crossings were named as ‘black spots’ (locations where motorists repeatedly violate
the road rules). Vehicles protruding into the path of the train and motorists entering
congested crossings and then panicking and driving backwards into the boom gates
were also mentioned.
283
Regional train drivers indicated that motorists not stopping or giving way to
trains is a continual problem at passively controlled crossings (i.e. no boom gates or
flashing lights). Regional train drivers generally agreed that the majority of motorists
obey protection systems; however some motorists drive through flashing lights or
drive around boom gates. Other high risk behaviours included motorists attempting to
beat the train across the crossing, speeding up to go through flashing lights, and
general risk taking by younger drivers in particular. Motorists not allowing enough
time to cross in front of the train or hesitating (stop-starting) at crossings were also
noted to be at high risk. There was a general perception by regional train drivers that
motorists are unable to judge the speed and distance of an approaching train to
determine a safe gap during which to cross. Local motorists were also reported to be
a problem at level crossings for regional train drivers. There was a general consensus
that motorists that are familiar with certain crossings hold low expectations of trains
being at a crossing and therefore are more likely to be complacent with scanning of
trains and the road rules. Interestingly, there was general agreement by the regional
group that about 60% of ‘near misses’ are due to poor visibility (such as angle of
approach) with only 30-40% actually due to the motorist. In the regional group,
motorist inattention and distraction were common issues identified, with the concepts
of inattentional blindness (not seeing the train coming) being a key concern
mentioned. However, the main cause of concern for both regional and metropolitan
train drivers was noted to be heavy vehicles.
A theme common to regional and metropolitan train drivers was the risk of
catastrophic consequence associated with level crossing collisions. The reasons given
for this were the threat of derailment, serious property damage, the high risk of a
fatality, personal injury and, most earnestly, the potential for enduring psychological
consequences. Drivers uniformly spoke about the continual fear they had of being
involved in a collision with a heavy vehicle, and many spoke of the effects that such
collisions had on train drivers involved. For this reason, train drivers were said to
consider any near-miss incident involving trucks particularly serious. This is in
marked contrast to the general view of near-crashes involving cars, which most train
drivers seem to rationalise and dismiss as a minor danger and acceptable part of their
job. Another emergent theme relating to heavy vehicle size and mass was the fact that
heavy vehicles have significantly more difficulty negotiating level crossings
284
successfully. This concern was reflected by the regional and metropolitan drivers in
different ways.
While regional train drivers were more inclined to discuss the difficulties heavy
vehicle drivers face in gauging the time required to cross the crossing safely, their
metropolitan counterparts cited concerns related to unintentional overhang of long
vehicles. This overhang was said to be especially prevalent in high traffic areas, where
a heavy vehicle driver underestimated the space required to clear a crossing, and were
thus unable to obtain safe clearance. Both the regional and metropolitan drivers
reached consensus that these risks often existed not because of dangerous behaviour
on the part of the heavy vehicle drivers, but inadvertently from a general ignorance as
to the actual size of their vehicle. Urban train drivers were further inclined to implicate
traffic, road design and crossing location as factors increasing this risk.
The consensus of train drivers regarding risk caused by unsafe heavy vehicle
motorist behaviour, was that it occurred frequently and willfully. A common
perception, especially among the regional train driver group, was that heavy vehicle
drivers often deliberately increased their speed in order to ‘beat the train’ across the
crossing. Attributions for this behaviour centered on a perceived desire to avoid delay.
While some drivers took the view that this behaviour was generated by impatience or
recklessness, others cited the intense time pressures that truck drivers are often placed
under. To this end, the drivers in the regional sample were able to cite that drivers
employed by two specific trucking companies were chief offenders. The wilful unsafe
behaviour by the heavy vehicle drivers emerged as a more serious problem for the
regional sample, with many examples of repeated violations at crossings suggesting it
to be of primary concern. However, train drivers within the metropolitan sample were
also able to identify variations of this behaviour, such as the deliberate avoidance of
boom gates.
For both the regional and metropolitan train driver focus groups, lack of
enforcement at level crossings also emerged as strong themes. Additionally, motorist
knowledge was also thought to be a contributing factor to behaviour with a general
consensus by both regional and metropolitan train drivers that there is low level of
knowledge of a train’s stopping distance, public misconceptions on the ability of the
train to stop and the poor understanding of the meaning of warning lights.
Furthermore, both regional and metropolitan train drivers reported that motorists in
general do not have the ability to judge the distance of a train or the speed of an
285
approaching train. Evidence of this perception was noted to be the frequency of near-
misses with vehicles at crossings.
7.2.2 Planning and development of interventions for each road user group
Study Two was undertaken as part of this research program involved formative
research as part of the planning, development and delivery of behavioural
interventions for each of the three road user groups identified in Study One. This
study also used both qualitative and quantitative data collection methods to provide
methodological triangulation and ensure reliability of the data. The overall objective
of the qualitative data collection was to obtain rich data using a qualitative mode of
inquiry, based on the key variables of attitudes, norms, self-efficacy (perceived
behavioural control), perceived risk, environmental constraints and the skills/abilities
of drivers. The overall objective of the quantitative data collection was to prioritise
the issues identified in order to direct and allocate project resources for intervention
planning, development and delivery. This combined recruitment strategy was adopted
as it was an appropriate and practical data collection strategy within the qualitative
and exploration methodology. Information obtained from each of the groups was
critical in assisting, guiding, and identifying priority areas for message and material
development.
7.2.2.1 Qualitative research
Qualitative formative research from each of the three road user groups revealed
that risk taking behaviours and attitudes differed greatly between the three road user
groups as well as between urban and regional settings.
For the older participants, most displayed high levels of knowledge and reported
low risk taking behaviour, with age-related factors acknowledged by the group as
being risk factors for their involvement in a collision. No older participant admitted
to intentionally violating or would be willing to violate the road rules at level
crossings such as driving through activated warning systems or failing to stop/give-
way to trains at passive crossings. Many participants mentioned the difficulty they
face with the ageing process and how it influences their driving ability. Decreased
vision, poorer hearing, slower reaction times and reflexes were the predominant
factors that were declared to pose problems when driving at level crossings. Poor
visibility at nighttime (due to degeneration of their vision) was also stated by many
286
participants as being a problem for driving at level crossings. This was a reason that
many participants stated that they avoid driving through crossings at night or when
there are low levels of light (such as dusk or dawn). Some participants mentioned that
their reflexes and flexibility were not the same as they used to be. Reflexes and
flexibility were felt to be an issue at crossings that have a sharp angle of approach
where participants have to turn to look back up the track. Many participants in this
group indicated they as they got older, driving generally became harder physically for
them. Quite a few participants also suggested that due to degenerating vision and
hearing that driving at level crossings was stressful for them and that they would
avoid situations that they were not confident driving.
Although both the urban and regional older groups indicated that they don’t take
risks at level crossings, many participants in the urban group mentioned the
difficulties that they sometimes face when driving in such complex traffic
environments (such as at busy urban level crossings). Behaviour of other motorists
was considered to be an issue with driving safely at level crossings for the older urban
group. Most participants felt that other motorists put pressure on them by beeping the
horn, yelling at them, or driving around their vehicle when they are waiting for
flashing lights to cease flashing. The majority of participants felt that such behaviours
were unnecessary and that younger drivers were largely responsible for such
behaviours.
The most notable differences between regional and urban older group responses
were related to risk perception and knowledge about the different types of protection
systems. Regional motorists appear to believe that they are at greater risk of being
involved in a level crossing collision with many recalling near-misses. The main
reasons given for such near-misses largely included complacency due to familiarity of
local level crossings. On the other hand, urban motorists generally held the
perception that level crossings were not as dangerous as other aspects of driving, with
many participants doubtful that motorists die at such intersections. Many participants
in the urban group also were surprised that transport authorities were investing
financial resources into level crossing education programs for motorists as the
majority believed that it is not a major road safety priority. Urban motorists also had
a poor level of knowledge about the different types of protection systems, other than
active systems with boom gates (such as in metropolitan areas), with some
287
participants unaware that some level crossings have no active protection system to
warn motorists of an approaching train.
According to the majority of participants in the older driver group, educational
strategies would best be focused on reminder or re-affirming messages as they believe
they always obey crossing rules and try to drive as safely as possible. Both regional
and urban participants felt that ‘informational’ television campaigns would be
effective for many participants in this group, especially when combined with visual
effects. Additional to television campaigns, the majority of participants indicated that
talk-back radio would be an effective delivery method as they frequently listen to the
radio and value the sources of information that talk-back radio provides. Public talks
from police, ambulance officers or train drivers were also seen as being valuable to
this age group. Hence, radio messages or public talks at social groups appeared to be
the most acceptable method of delivery for educational campaigns for the majority of
participants in the older group.
Younger drivers demonstrated low risk perception of the consequences to
unsafe driving behaviour at level crossings, with risk taking being reported at high
levels for this group. Participants that spoke of performing high-risk behaviours (such
driving through flashing lights whilst the boom gate is descending) indicated that they
felt that their chances of being hit by a train were small. The majority of these
participants that admitted to such risk taking behaviours indicated that this type of risk
taking was less dangerous than other risk taking behaviours such as drink driving,
speeding or driving through red lights at an intersection. Some participants stated that
they frequently disobey the rules at level crossings and have never had a near-miss
with a train. The reasons for such justifications included: infrequent trains, being able
to judge the distance that the train is from the level crossing, and that trains will slow
down if they observe a car at a crossing. Those participants that perceived there was
some degree of risk of being involved in a collision, thought it would be due to either
equipment malfunction (such as faulty boom gates/flashing lights) or stalling their car
at the crossing. Many participants in this group raised the concern that if they stalled
their vehicle on the crossing then this would be extremely dangerous. Additionally,
the majority of this group believed that at crossings with low train traffic volumes
they are less likely to be involved in a collision than at crossings with more train
traffic volumes. Male participants were more likely to perceive that being familiar
with certain crossings would make them less likely to be involved in a collision with a
288
train, while female participants were more likely to hold the perception that driving at
crossings on a regular basis would make them more likely to be involved in a collision
with a train due to being on ‘auto-pilot’. Peers as passengers were also seen as a key
element in driving safely at level crossings, with many participants stating they would
take more risks when driving by themselves. The risk of damaging their car was a big
influence on whether younger driver participants drove safely at crossings or not.
Some participants stated their car was very important to them and if they were
involved in a collision with a train (and not injured) then they would be distressed if
their vehicle was destroyed as they had ‘saved up’ their money to purchase it.
Overall, there was very little difference between regional and urban group
responses. Both groups believe that they are at low risk of being involved in a
collision with a train as they believe that they have the skills and abilities to avoid a
collision. Interestingly though, some participants in both groups indicated that they
find it difficult to judge the distance a train is from a crossing which appears to
oppose such views of being able to avoid a collision. Impatience for waiting at
crossings was evident in both the urban and regional groups, with regional drivers
being frustrated with the time it takes for freight trains to pass while urban drivers
were frustrated with waiting for protection systems (i.e. boom gates or flashing lights)
to turn off. The regional group reported a high acceptance of a range of risk taking
behaviours (‘beat the train’), while participants in the urban group tended to report
lower risk taking behaviours (poor knowledge of road rules). Of particular concern
with both the regional and urban younger drivers was their willingness to take risks.
Like the inconsistency in both group’s perceived level of risk at level crossings, there
appeared to be a paradox in self-reported driving behaviour. Most participants in both
groups stated that they frequently take risks at crossings (such as driving through
flashing lights or when a boom gate is descending), however the majority of
participants mentioned that other motorists caused them to take such risks.
Nearly all of the younger participants thought that mass media campaigns were
effective in changing their and others behaviour. The most acceptable method of
delivering educational interventions for this group appears to be through radio or
television. Some participants stated that they thought shock television campaigns were
the best for young drivers while other participants indicated that they would just tune
out to shock style campaigns. A scenario involving risking their friends’ lives when
driving unsafely was a common theme for developing messages.
289
Heavy vehicle drivers indicated a high level of knowledge of safety issues, with
engineering and design being a major factor contributing to near-misses or the risk of
being involved in a collision at a level crossing. A unanimous complaint from heavy
vehicle drivers was that level crossings are not designed in a manner which is user-
friendly to their vehicles. The metropolitan environment was cited as a primary
example, where choice of crossing location, traffic and other roadway factors can lead
to unintentional short-stacking and overhang (e.g., required to stop while part of
vehicle remains on crossing). On the other hand, regional crossings were also said to
cause difficulties, with design faults and location choice having detrimental effects on
sight distances and train visibility. A further comment was inadequate warning of
approaching crossings, which drivers perceived greatly reduced their ability to take
appropriate action. Heavy vehicle drivers suggested that such shortcomings in design
and protection systems were responsible for the majority of unsafe driver behaviour
observed at level crossings.
However, a high level of familiarity with their routes and with the crossings
they encountered while traveling these routes, was also nominated by heavy vehicle
drivers as a major factor associated with their crossing behaviour. Here, it was discussed
that drivers sometimes suffered lapses of concentration, resulting in them failing to follow
appropriate safe crossing behaviour at level crossings. Drivers tended to defend their own
perceived complacency as a function of high levels of familiarity with the crossings and,
furthermore, expressed a degree of confidence in their ability to identify which crossings
were most dangerous and required most attention. While a practical approach on the
surface, this familiarity was sometimes observed to breed overconfidence in their abilities
to control their vehicle, which resulted in the drivers being generally less inclined to
engage in safe crossing behaviours.
Though the heavy vehicle drivers were generally inclined to identify the above
two factors as the predominant causative agents of their unsafe crossing behaviour, a
minority raised the spectre of wilful risk taking among drivers. These reports included
both confessions of individual involvement, as well as second-hand reports from and
observations of other drivers’ behaviour. The most prominent motive for unsafe
behaviour was a desire to avoid delay in getting to their destination. These drivers
frequently cited the delay caused by not only waiting for trains to pass, but also the
significant time lost involved in deceleration and re-acceleration. Others were more
direct, citing the personal frustration and impatience associated with stopping at
290
crossings for professional drivers. These participants stated that trains should be
slowed down at crossings with large volumes of heavy vehicles.
At the core of each of these behaviours were perceptions of time pressures due
to the rigid timetabling imposed by the trucking companies. Many drivers attempted
to rationalise their behaviour by stating that, although they knew not obeying level
crossing warning systems to be dangerous, they managed their risk level by limiting
this behaviour to crossings where they perceived such behaviours to be ‘safe’. It is
worth noting that drivers were, in general, uncomfortable when discussing wilful risk
taking, often attributing it to ‘other drivers’. This points to a general awareness that this
behaviour is not perceived as responsible, safe or socially acceptable.
There were mixed opinions about educational interventions targeted towards
heavy vehicle drivers, with most drivers indicating that the most acceptable methods
would be in sharing of personal experiences (through trucking magazines), awareness
raising (through company training sessions) or radio messages.
7.2.2.2 Quantitative research
For the quantitative component of Study Two’s formative research, both train
drivers and a panel of experts completed questionnaires to ascertain differences in
beliefs of high risk behaviour by motorists at level crossings. This 19-item
questionnaire (using Likert scales) was developed from results obtained from the
qualitative formative research conducted with each of the three road user groups.
Themes that emerged from this qualitative research (i.e. high risk behaviours of
drivers) were selected by a panel of three road safety experts (separate to those
experts participating in this survey). An independent t-test was conducted to compare
overall mean scores between the two groups: train drivers and experts. There was no
significant difference between overall mean scores. However, both experts and train
drivers ranked the items driving around the boom gates and trying to beat the train
across the crossing as being the highest risk taking behaviours. Experts also ranked
overtaking cars stopped at the crossing as being very high risk, while train drivers
ranked drive in front of a train when it is close to the crossing as being a very high
risk taking behaviour. Overall, experts were observed to rank the behaviours lower
than train drivers. This may be due to the nature of train drivers observing a large
number of near-misses throughout their careers and the possibility that they
291
experience great frustration with motorists that take risks at level crossings (findings
from the focus groups discussions with train drivers in both urban and rural areas).
7.2.3 The present context of motorist behaviour
There were three parts to the final study (Study Three). The aim of Part One of
this study was to develop targeted interventions specific to each of the three road user
groups using Fishbein’s model (Integrated Model of Behaviour Change). Part Two
involved the investigation of the present context of unsafe driving behaviour at level
crossings. This included examining the personal, social and environmental factors
contributing to unsafe driving intention and driving behaviour at level crossings. Part
Three involved trialing a pilot road safety radio advertisement using an intervention
and control methodology. A summary of the findings from Part Two and Part Three
are presented below.
7.2.3.1 Heavy vehicles at level crossings
Research undertaken with heavy vehicle drivers in Study Three suggested that
truck drivers hold a number of beliefs in relation to other motorists and environmental
constraints at level crossings. Firstly, heavy vehicle participants indicated that other
motorists drive unsafely more frequently than their colleagues or friends. This finding
is not surprising, given the comradeship that is known to exist amongst truck drivers.
Secondly, the majority of participants believe that they will never be involved in a
level crossing collision, although some indicated that it is acceptable to sometimes
disobey road rules at level crossings. Given that level crossing collisions are rare
events, it is not surprising that this small sample would have a low perception of risk
at level crossings. What is surprising is that participants believe that it is generally
not possible to judge a train’s speed, but that it is still safe to cross when warning
systems indicate that there is a train approaching the crossings. Two environmental
(or engineering) factors were considered to be the major cause of unsafe crossing
behaviour by truck drivers: blinding sun making it difficult to see if flashing lights are
activated and the design of the approaching road (such as an ‘S’ bend) makes it
difficult to see if a train is approaching a crossing. These two issues were a recurring
theme that was raised by truck drivers at different stages of the research program.
292
The most common solution raised by participants is advanced warning systems (on
the road approaching the level crossing).
Truck drivers that work shifts were found to be significantly more likely to take
risks when driving generally on the road and also at level crossings. These shift
workers were also considerably more likely to hold the belief that environmental
constraints affect their ability to drive safely at level crossings. Given that flexible
approaches to prescriptive driving hours regulations may lead to improved road safety
outcomes (Mabbott, 2001), one would expect that reducing shift workers exposure to
level crossings through route changes, may have a long-term effect on level crossings
collision rates for the heavy vehicle industry. Familiarity was also a key issue with
driving a truck at level crossings. Participants that were familiar with driving at
crossings with only flashing lights recorded higher levels of unsafe driving at level
crossings generally, slightly higher risk taking driving behaviour and higher levels of
self-efficacy than unfamiliar drivers. The current study gained information relating to
self-reported driving behaviour at level crossings as well as self-reported exposure
(i.e. how frequently a participant crosses a particular protection system). The purpose
of obtaining such data was to gain information on exposure to the three types of
protection systems at level crossings. The finding that familiarity is associated with
higher levels of risk taking has significant implications for the heavy vehicle industry
given the potential for a major catastrophic event to occur if a truck collides with a
train.
As can be seen, there are a number of findings from this research that have
implications for improving level crossing safety for both heavy vehicle drivers and the
greater public. However, the nature of professional truck driving (i.e. commercial
practices and arrangements) poses some significant obstacles to achieving improved
safety outcomes. Firstly, with the increasing adoption of ‘just-in-time’ ware housing
polices comes an increase in the importance of the express freight sector, as well as
both overnight distribution and long distance transport (George, 2002a). Secondly, in
recent years that has been significant growth in specialist vehicle types (such as B-
doubles) which has changed the mix of heavy vehicles on the road as well as a higher
proportion of heavy vehicles in the traffic stream (George, 2002a). The combination
of these two factors weighs heavily on whether or not safety can be maintained, let
alone improved, at level crossings.
293
7.2.3.2 Older drivers at level crossings
There were a number of interesting findings from the older driver sample in
Study Three. Older drivers were asked a number of questions about their health and
driving ability. Of greatest concern is that many participants reported substantial
decline in their hearing, restricted range of motion to their neck and having increasing
trouble adjusting to glare and night-time driving. These factors have significant
implications for older drivers (60 plus years) traversing level crossings. Firstly,
hearing warning systems and trains approaching a crossings (passive crossings in
particular), is fundamental in ensuring safe driving behaviour. Secondly, with
restricted movement to the neck whilst driving, adequate scanning for trains at
crossings that have approach angle that is oblique (i.e. an angle, such as an acute or
obtuse angle, that is not a right angle or a multiple of a right angle) would be sub-
optimal for safely negotiating such level crossings. Thirdly, have difficulty with glare
when driving may have an affect on the ability of older drivers to accurately observe
if flashing lights are activated at crossings without boom gates. These three factors
may very well play a critical role in the over-representation of older drivers in level
crossing collisions.
Self-efficacy was shown to be central for older drivers at level crossings. Older
drivers reported lower levels of self-efficacy of driving at level crossings as the
majority believe that other motorists influence their driving and their ability to obey
the road rules. Such findings however contradict previous research into speeding.
Elliot et al (Elliott, 2003b) in their research found that older drivers had greater
perceived behavioural control (perceived self-efficacy) then younger drivers when it
came to complying with speed limits. Within the current study, anecdotal evidence
from experts in the field (Study One) revealed that older drivers may feel pressured by
other motorists to take risks at level crossings. The beliefs held by experts in Study
One were observed to have support by the findings of Study Three. Subjective norms
of other motorists driving were also found to score higher for older drivers. Older
driver participants hold the belief that other motorists do not obey the rules at level
crossings as often as their important others (i.e. family and friends). This finding
supports the formative research undertaken with older drivers in that they hold the
belief that their age group drives more safely at level crossings than other road user
groups. Also of interest is the poor level of knowledge of the road rules at level
294
crossings. Anecdotal evidence had suggested that many motorists are not familiar
with the road rules at level crossings; this older driver sample certainly supports this.
Like the heavy vehicle participants, familiarity with driving at level crossings
was found to be a factor related to unsafe driving of older drivers both generally and
at level crossings. Older drivers that were familiar with driving at crossings with
boom gates recorded higher levels of unsafe driving generally than unfamiliar drivers.
Additionally, it was found that older drivers that were familiar with driving at
crossings with only flashing lights recorded higher levels of unsafe driving and
intended driving behaviour at level crossings than unfamiliar drivers. Furthermore,
the study found that older participants that are familiar with driving at passive
crossings are more likely to have a negative attitude towards driving at level crossings
generally. It appears that many older drivers use self-protective behaviours such as
avoiding driving at level crossings during times they perceive may be ‘riskier’ to them
(i.e. dusk, peak traffic periods). This process of self-regulation has been suggested to
be related to deterioration in vision, general health and crash involvement (Ragland et
al., 2004). It could be argued that this self-regulation also involves using roads that
older drivers are familiar with; including level crossings. Interventions such as
awareness raising of how familiarity may play a role in crashes at level crossings for
older drivers, may assist in discontinuing the current over-representation of this road
user group in level crossing collisions.
7.2.3.3 Younger drivers at level crossings
There are a number of interesting findings from the younger driver sample.
Firstly, compared to older and heavy vehicle participants, younger participants scored
substantially lower in their knowledge of road rules at level crossings. This is an
interesting finding as younger drivers would most likely have been exposed to some
information about level crossing road rules in the licensing process. Secondly,
younger drivers reported higher levels of risk perception (i.e. involvement in a
collision with a train at a level crossing) compared to older and heavy vehicle drivers.
Historically risk analysts propose that the most accurate measurement of driving risk
is collision rates, however using such rates have limitations for level crossings. Even
though collision rates at level crossings are substantially lower than that of the
national road toll, it does not necessarily mean that the risk is not high (Wang, 2002).
295
Wang (2002) proposes that a high risk could have been perceived but avoided. This
finding was also surprising given that research has found that drivers that believe they
are less vulnerable to crashes than other drivers, have less incentive to engage in self-
protective behaviours (McKenna, 1993, Horswill et al., 2004). It would appear that
the findings from the younger driver sample contradict findings from this previous
research. One explanation for this is that although younger drivers hold the
perception that level crossings are high risk intersections, they are less likely to be
fatally injured at a level crossing than from speeding or other high risk driving
behaviours. Another explanation could be their low exposure to level crossing driving
compared to their participation in other unsafe driving behaviours.
The third remarkable finding is that related to self-efficacy. Self-efficacy has
been defined as an individual’s capacity to organise, control and execute certain
behaviours to attain specific performances (Bandura, 1977a). However, since self-
efficacy is regarded as a coderterminant of behaviour (Ajzen, 1991), it relies on the
accuracy of an individual’s perception of their control over a driving situation
(Sheeran, 2003). For the current younger driver sample, self-efficacy was found to be
a predictor of intention to drive safely at level crossings. As such, lower levels of
self-efficacy were associated with a stronger intention to drive unsafely at level
crossings. The current research produced similar results to previous research with
younger drivers. Elliot et al (2003b) in their research about driver’s compliance with
speed limits found that younger drivers (and particularly males) had lower levels of
self-efficacy than older drivers. Some research has indicated that individuals with
higher levels of self-efficacy report significantly greater rates of success in changing
behaviour (Wells-Parker et al., 2000). However, one of the major assumptions of
Fishbein’s Integrated Model (IM) is that attitudes, norms and self-efficacy (as well as
their underlying beliefs) have primarily an indirect effect on behaviour (via their
effects on intention) (Fishbein et al., 2003). Thus, if young drivers have already
formed strong intentions to perform unsafe driving at level crossings (but are not
doing so), then little will be accomplished by developing interventions designed to
increase self-efficacy (Fishbein et al., 2003). This has important implications for
designing educational programs for young drivers. Since the majority of younger
drivers appear to hold intentions to drive safely at level crossings, future interventions
should be directed at increasing relevant skills and abilities and/or assisting with
environmental constraints.
296
Another noteworthy finding of this study with younger drivers is that young
male participants were significantly more likely to report both current and future
driving at level crossings to be less safe than their female counterparts. This is not
surprising given the amount of literature supporting research that young males are
more likely to take risks whilst driving generally (Laapotti and Keskinen, 1998,
Arnett, 2002) or be less ‘safety-orientated’ than their female counterparts (Meadows
and Stradling, 1999). However, this finding has implications for targeting younger
drivers as it appears that educational programs may need to direct efforts largely
towards young male drivers as opposed to their female counterparts. The challenge
remains for program designers to implement interventions that have the prospect to
produce lasting behaviour change in young males.
Finally, the finding that familiarity with driving at level crossings was
associated with unsafe driving was not surprising given that this result was seen in
both the older and heavy vehicle samples. What is surprising is that younger drivers
are most likely to be less familiar with different designs of level crossings than the
older and heavy vehicle participants, yet they report more unsafe driving behaviours
compared to these other two samples.
7.3 IMPLICATIONS FOR LEVEL CROSSING SAFETY
Previous research into unsafe driving at level crossings supports the notion that
from a systematic point of view, level crossing collisions are far more complex than
the immediate causal factor being the non-observance of the road rules (Caird, 2002).
Although the act of driving is taken for granted by most motorists, it is however a
complex task that is dependent on a variety of cognitive and psychomotor
performance abilities to be intact, such as alertness, attention, multitasking, memory,
coordination, and visual spatial perception (Moller, 2004). As level crossings are the
interface between the rail and road, and are considered to require similar cognitive
processes to road intersections, they pose a unique problem for drivers. Driving
through intersections is possibly one of the most complex conditions drivers
encounter, as multifarious perceptions, decisions and maneuvers are required to
successful negotiate and cross intersections (Trbovich and Harbluk, 2003). This
background provided the foundation for the three studies conducted in this research.
297
In the course of conducting this research, a number of significant implications for
managing risk at level crossings have been identified.
Firstly, it has been shown that older drivers hold low levels of self-efficacy for
driving at level crossings, with other motorists influencing their ability to safely
negotiate crossings. If this perception of driving control at level crossings is
influenced by older driver’s functional, perceptual or cognitive impairments, remains
unknown.
What this research has highlighted is the prevalence of older driver impairments
in some important functional abilities that may be directly related to level crossing
driving: increasing trouble adjusting to glare and night-time driving, restricted range
of motion to the neck and substantial declines in hearing. Although factors
contributing to the over-representation of older drivers in collisions at level crossings
are likely to be complex and multi-faceted, such functional impairments are expected
to play a critical role. So too is the ability of older drivers to adopt safe driving
practices by self-regulating their driving behaviour. The connection between
familiarity and unsafe driving practices for older drivers, also poses another problem
for ensuring safety for this age group. Some research has indicated that many older
drivers are aware of what generally constitutes a risky driving situation, yet are
unaware that some driving conditions (such as complex intersections) have become
specifically risky for them (Holland and Rabbitt, 1992). This notion may hold true for
older drivers at level crossings. As such, they may over-estimate their own driving
ability while under-estimating their risk of being involved in a vehicle-train collision.
Secondly, younger drivers recognise that level crossings are potentially a highly
dangerous intersection, albeit that collisions at crossings are rare events, yet are still
likely to engage in risk taking behaviours. Additionally, their low levels of self-
efficacy in driving at level crossings pose challenges for developing interventions
with this age group. If the success of changing behaviour is related to increasing
driver’s levels of self-efficacy (Wells-Parker et al., 2000), then future interventions
are more likely to be effective if they are directed at increasing relevant skills and/or
abilities as well as assisting with environmental constraints. Given that experience
constitutes the core of skills and abilities of drivers, and the lack of financial resources
to actively protect younger drivers at any level crossing they encounter, how such
interventions are applied in the ‘real-world’ is likely to be an arduous task. It would
298
appear that the licensing process is one critical element in preparing younger drivers
for the multitude of hazardous driving scenarios that exist in the driving environment.
Thirdly, most heavy vehicle drivers reported driving safely and intending to
drive safely in the future. However, there is a sub-set of drivers that indicate they
have in the past and will in the future take risks when traversing crossings. Although
this sub-set is relatively small, if generalised to the larger trucking industry it could be
problematic for the rail sector and greater public alike. The fact remains that heavy
vehicle-train collisions, although rare events, have the potential to be catastrophic in
terms of fatalities and injuries, environmental disaster, delays in the rail network, and
extensive damage to property. In June 2007, such a catastrophic disaster was
observed in northern Victoria when a train was derailed killing 11 train passengers.
With this tragic outcome came major media attention and great negativity towards the
trucking industry. Nevertheless, such events are likely to continue with the
commercial traffic on the eastern corridor of Australia forecast to double in the next
5-10 years, and the complex set of social practices associated with professional truck
driving placing pressure on truck drivers. The two greatest pressures are allied with
the increasing adoption of ‘just-in-time’ ware housing (increasing the importance of
the express freight sector, as well as both overnight distribution and long distance
transport) and the significant growth in specialist vehicle types (i.e. B-doubles) which
drastically changes the mix of heavy vehicles on the road as well as a higher
proportion of heavy vehicles in the traffic stream (George, 2002a).
The ultimate goal to improve level crossing safety for all motorists would be to
have a combination of engineering, education and enforcement countermeasures for
all level crossings. However, the small number of fatalities in comparison to the
national road toll limits this. To date, there’s been very little enforcement at
Australia’s 9400 level crossings and police have indicated their resources will
continue to be largely concentrated on speeding, drink/drug driving, fatigue and
seatbelt usage. Unless police ministers around the country change policies to increase
police presence at ‘dangerous’ crossings, the role that enforcement practices takes will
remain the same. Nevertheless, the likelihood of creating behavioural change would
be increased if violations at level crossings by all motorists was detected and
penalised, or alternatively, if perceptions of such detection were increased. Since
deterrence is achieved through instilling fear in drivers with the threat of punishment
via some form of sanction (Elliott, 2003a), it is aimed mostly at the ‘marginal group’
299
who lie between the law abiding citizens (for who deterrence is not necessary) and the
undeterrable citizens (for who deterrence is ineffective) (Zimring and Hawkins,
1976). Further research that examines the appropriateness of deterrence theory may
provide another perspective for which to tackle level crossing safety.
As described previously, there exists in Australia a comprehensive system for
assessing risks at level crossings and the likely impact of different treatments on these
risks. After determining a level crossing is high risk and what makes up the risk, the
system is used to determine treatments appropriate to reduce the risk. This
assessment system has the capacity to assess benefit/cost for proposed improvement
works, ensuring that each dollar is spent where it can generate the greatest safety
improvement. Unless significant financial resources are bestowed by both
Commonwealth and State governments to upgrade all level crossings to actively
protect road users, the current assessment system will continue to allocate the limited
financial resources to the nominated crossings. Although making wide-spread
changes to current operational level crossings may be considered unfeasible, future
level crossing design (including upgrade programs) should accommodate for the
changing mix of trucks in the traffic stream. More specifically, the results of this
research indicate that the level crossing design process should include consideration
of the requirements and limitations of driving large vehicles, including: (a) length of
vehicle, (b) maneuverability, and (c) visibility. Research across the Canadian rail
network, comparable to Australia’s in terms of its size and layout, supports such
findings, having demonstrated that the design of level crossings often do not
accommodate for the specific needs of heavy vehicles (Gou and Bellvigna-Ladoux,
2003). Furthermore, the implementation of advanced warning systems such as
increasing the existence of early road signage may prove fruitful in improving safety.
In addition to enforcement and design factors, an educational awareness
campaign at the national level that is viewed by both metropolitan and rural drivers
would highlight the risks of driving for different age groups. However, such
campaigns must be developed and directed specifically to road user groups (i.e.
younger drivers, older drivers etc) if they are to have any substantial effect on
improving safety for all motorists. For the trucking industry, there may appear some
merit in implementing educational awareness campaigns that highlight to large
vehicle drivers the increased risks associated with operating trucks on and around
level crossing. Such an initiative could be conducted at a company level through the
300
supervision or induction process and possibly reinforced through corresponding
regulatory governing bodies. The output of this exploratory research can be
conceptualised as the ‘groundwork’ required in building specific, targeted road safety
educational countermeasures for different road user groups.
7.4 STRENGTHS AND WEAKNESSES OF THE RESEARCH
The strength of this research program is found in the contribution of knowledge
about the influencing factors on motorist safety at level crossings. This research
program is the first of its kind in the world. Although interventions have been
developed in other countries, none have been guided by empirical research or theory.
The motivation underlying motorist behaviour determines to a large extent how
successful behaviour change strategies may be. This research program has been
guided by Fishbein’s Integrated Model of Behaviour Change (IM) to assist in the
planning and development of interventions as well as guiding instrument development
to ascertain the present context of motorist behaviour at level crossings.
What makes this research unique is its inclusion of train driver’s perspective
with those of three distinct road user groups (i.e. older and younger drivers, and heavy
vehicle drivers). In doing so, this research has added to the existing literature as well
as providing greater insight into why motorists take risks at level crossings.
Furthermore, a multi-method approach to investigating factors that influence unsafe
driving behaviour at level crossings allowed for a more thorough examination of the
issues. The method of ‘triangulation’ (i.e. combining research methods to give a range
of perspectives) whereby both qualitative and quantitative research designs were
utilised, provided the robustness required within the discipline of road safety research.
In addition to the strengths of this research program, there are weaknesses. One
limitation of this research is the low response rate among heavy vehicle drivers. This
response rate resulted in a small sample size that potentially limits the validity and
generalisability of the findings, as well as the range of statistical analyses that could
be undertaken. The reluctance by many trucking companies to participate in this
research may have stemmed from either the belief that level crossing safety is not
high on their safety agenda or a distrust of disclosing driving behaviours of employees
in fear of scrutiny of the company’s safety compliance. As such, this study may have
only captured trucking companies with comprehensive OHS management systems and
301
a focus on ensuring a culture of safety within their workplace. Additionally, the high
attrition rate observed for the heavy vehicle sample may have been partly attributable
to the length and number of questionnaires. Moreover, this small sample size also
prevented randmonisation of companies to either the intervention or control group in
Study Three from occurring. This was a major limitation for this road user group and
further techniques and incentives for company participant may need to be developed
in order to increase the industry’s participation rates in road safety research over
longer periods of time.
This investigation also suffers from the typical perceived limitation associated
with self-report data. The main criticism of self-report data is that it is affected by
social desirability bias. It goes without saying in road safety research, that social
desirability will have some impact upon the findings obtained. It has been suggested
that ‘deviant’ individuals would minimise the number of crashes they were involved
in or other illegal driving activities (e.g. drink driving) (Hatakka, 1997). However, it
is important to note that any resulting associations found would be under-estimates of
real associations, rather than over-estimates (Hatakka, 1997, Lawton, 1997a, West,
1995). For the current investigation, the use of the self-report data would have
potentially been a problem if actual driving behaviour was monitored. However,
since actual driving behaviour was not monitored, the degree of correspondence
between self-report and observed driving behaviour is most likely to be insignificant.
Since there exists no standardised measures of attitudes, norms, self-efficacy,
driving intention, perceived risk, and environmental constraints for level crossing
behaviour, the researcher was required to develop scales to directly and indirectly
measure such constructs. This may have affected the validity and reliability of the
instruments. Another widespread limitation of this investigation is that data was
drawn exclusively from only one jurisdiction in Australia. Every effort was made to
ensure that there was a diverse sample of participants with an equal number of
recruitment letters being sent to rural/remote and metropolitan drivers. However, with
financial constraints limiting the scope of data collection extending outside of
Queensland, caution needs to be exercised when generalising these results to other
jurisdictions in Australia or abroad.
Finally, although this research has found some important factors that may
influence unsafe driving by motorists at level crossings, explaining the causal
mechanism is complicated. From a systematic point of view, level crossing collisions
302
are far more complex than the immediate causal factor being the non-observance of
the road rules (Caird, 2002). The exploratory nature of this research program limits
conclusions and generalisations being made. As such, similar investigations are
needed in examining these road user groups before these findings can be exclusively
applied at the macro level. Suggestions for further research are provided in the
subsequent section.
7.5 SUGGESTIONS FOR FUTURE RESEARCH
Although the multi-method approach to investigating factors that influence
unsafe driving behaviour at level crossings allowed for a more thorough examination
of the issues, further qualitative research would be useful to clarify and expand on
some findings of this research program. These include how truck drivers can reduce
their risk of being involved in a catastrophic event if their truck collides with a train.
Even though anecdotal evidence indicated that road safety interventions which
are brief in nature (i.e. one-off interventions) are not likely to be highly effective, no
research in level crossing safety has ever been conducted that has examined such
interventions. Due to the limited financial resources that both federal and state
governments currently allocate to improving level crossing safety, efforts to intervene
on specific road user groups through a single intervention educational approach
deserved consideration. Although this research concluded that ‘one-off’ brief
interventions are ineffective for changing driver intention for level crossings, the
output of this exploratory research can be conceptualised as the ‘groundwork’.
Considering that current mass media campaigns for level crossing safety do not
include any empirically guided interventions, this research is an important step for the
rail industry in ensuring that mass media campaigns are both informed by road safety
theory as well as the motivations underlying driving behaviour of the road user groups
(target audience) for which they are designed to target.
However, it must be kept in mind that with motorist error being considered an
important factor in the causation of level crossing collisions, reliance solely on
education is not the solution. How rail and road authorities combine the mix of
education, enforcement and engineering solutions to improve level crossing safety is
and will continue to be an arduous task.
303
7.6 RECOMMENDATIONS TO INDUSTRY
Based on the findings from this research program, there are a number of key
recommendations for improving level crossing safety.
The first of the key recommendations relates to the reporting of near-miss
incidents at level crossings by train drivers. Currently, it appears that it is up to a train
driver’s discretion whether or not they report near-miss incidents. The formative
research undertaken indicated that there exists a culture among train drivers that they
are ‘whingers’ if they report all near-miss incidents at level crossings. Many train
drivers indicated that the reporting mechanism for such events (i.e. transmission over
the radio to control) does not allow any anonymity and thus they are reluctant to
report the numerous incidents that they encounter. Additionally, train drivers hold the
belief that even if they do report such incidents, the likelihood of any action taken by
police is extremely low. Due to the large number of near-misses at level crossings
that do not involve a fatality but have the potential to result in a catastrophic event
such as a train derailment, it is important to identify the occurrences of near-misses as
well as collisions involving fatalities. As such, a key recommendation is for the
review of the current system for reporting of near-miss incidents to maximise the
accuracy of such events.
The second key recommendation relates to the recording of national collision
data. Findings from a review conducted by Ford and Matthews (2002) found that
most jurisdictions in Australia have “widely differing methods of categorising and
recording the level crossing characteristics and accident data” (p10). As a
consequence of the differing methods of recording of level crossing collisions in
Australia, there is a lack of definitive evidence available relating to the extent and
nature of level crossing collisions. Coupled with the inaccuracy of near-miss
incidents from low levels of reporting by train drivers, it is difficult to determine how
best to direct countermeasure resources. Cairney (2002) recommends that the
usefulness of information from collisions at level crossings would be greatly
improved by including the variables that are currently collected for road crashes, such
as vehicle type and driver characteristics.
Level crossing collisions have been shown to result in enormous human and
financial cost to society (Lobb, 2001). Although there have been attempts to
determine level crossing collision costs, the accuracy of such reports have been
304
revoked by both railway experts and insurance companies alike. It has been
suggested that estimations are grossly conservative. The most recent report assessing
the socioeconomic costs of vehicle-train collisions was conducted by the Bureau of
Transport and Regional Economics (2002). This analysis covered collisions that
occurred in Australia only during 1999. This report concluded that the total cost of
level crossing collisions was $32 million, with approximately $10 million related to
vehicle-train collisions (Bureau of Transport and Regional Economics., 2002). It is
recognised that costs of level crossing collisions cannot be quantified exactly,
however in order to receive more appropriate levels of funding from government
agencies to invest in improving safety and developing effective countermeasures, a
report that more accurately determines the direct and indirect costs needs to be
undertaken.
The design of crossings and absence of advanced warning systems appear to be
major issues for motorists, particularly heavy vehicles. As described previously, there
exists in Australia a comprehensive system for assessing risks at level crossings and
the likely impact of different treatments on these risks. After determining a level
crossing is high risk and what makes up the risk, the system is used to determine
treatments appropriate to reduce the risk. This assessment system has the capacity to
assess benefit/cost for proposed improvement works, ensuring that each dollar is spent
where it can generate the greatest safety improvement. Although making wide-spread
changes to current operational level crossings may be considered financially
unfeasible, future level crossing design (including upgrade programs) should
accommodate the changing mix of trucks in the traffic stream. It is evident from this
research that heavy vehicle drivers hold the belief that they experience significant
difficulty in negotiating crossings with an ‘S’ bend approaching road. Advanced
warning systems may assist drivers in preparing for a train at these crossing when
visibility on the approach road is reduced. As such, further investigation of this
matter needs to be considered.
Another strong recommendation that has emerged from this research is the need
to develop educational and awareness materials specific to different road user groups.
Although the interventions developed are one example of interventions for these road
user groups, the output of this exploratory research can be conceptualised as the
‘groundwork’ required in building specific, targeted road safety educational
countermeasures. The research found that there are unique issues that motivate
305
different road user groups. However, there is one common element for all road user
groups: familiarity. Familiarity with driving at level crossings was found to be
significantly associated with unsafe driving intentions for all road user groups. This
factor therefore should be included in future educational and awareness raising
campaigns as it is a critical element that has been found to be related to fatal level
crossing collisions (Wigglesworth, 1979). Additionally, it is advised that future
campaigns be developed and directed specifically to distinct road user groups if they
are to have any substantial effect on improving safety for all motorists.
For the trucking industry, there may be some merit in implementing educational
awareness campaigns that highlight to heavy vehicle drivers the increased risks
associated with operating trucks on and around level crossing. Such an initiative could
be conducted at a company level through the supervision or induction process and
possibly reinforced through corresponding regulatory governing bodies or
associations. Messages could also be delivered through trucking magazines as this
medium may indeed provide the most appropriate source for message acceptance in
this group.
The development of education and awareness materials for the older driver
group would need to be substantially different. What this research has confirmed is
the high prevalence of functional impairments for drivers that are 60 plus, such as
increasing trouble adjusting to glare and night-time driving, restricted range of motion
to their neck and substantial declines in their hearing. Although factors contributing
to the over-representation of older drivers in collisions at level crossings are likely to
be complex and multi-faceted, such functional impairments are expected to play a
critical role. Self-regulation for older drivers appears to play role in their driving at
level crossings and other busy intersections. Although many older drivers may be
aware of what generally constitutes a risky driving situation, they may be unaware
that some driving conditions (such as complex intersections) have become specifically
risky for them (Holland and Rabbitt, 1992). This notion may hold true for older
drivers at level crossings. As such, they may over-estimate their own driving ability
while under-estimating their risk of being involved in a vehicle-train collision.
Including such information in education and awareness raising materials may be
beneficial in reaching the older driver group.
Younger driver’s low levels of self-efficacy while driving at level crossings
pose challenges for developing educational and awareness raising materials for the
306
17-24 year age group. If the success of changing behaviour is related to increasing
driver’s levels of self-efficacy (Wells-Parker et al., 2000), then future interventions
are more likely to be effective if they are directed at increasing relevant skills and/or
abilities as well as assisting with environmental constraints. Given that experience
constitutes the core of skills and abilities of drivers, and the lack of financial resources
to actively protect younger drivers at any level crossing they encounter, how such
interventions are applied in the ‘real-world’ is likely to be an arduous task. It would
appear that the licensing process is one critical element in preparing younger drivers
for the multitude of hazardous driving scenarios that exist in the driving environment.
Another is most likely to be through driving schools. Driving instructors have a
“critical role in providing high-quality learning opportunities for novice drivers”
(Bailey, 2003,p1). Inclusion of mandatory level crossing driving for learner drivers
may provide an opportunity to ensure young drivers are exposed to high risk
intersections such as level crossings.
Coupling enforcement programs with educational and awareness raising
campaigns will no doubt provide some improvement in driver behaviour at level
crossings. How this can be achieved is a continuous challenge. Nevertheless, the
likelihood of creating behavioural change would be increased if violations at level
crossings by all motorists was detected and penalised, or alternatively, if perceptions
of such detection were increased. Increasing current fines/penalties and developing a
uniform national approach to offences at level crossings, is the preliminary step
towards deterring motorists in violating the road rules. Only recently in Queensland
and Victoria has there been a substantial increase in fines and penalties, with the
catalyst in Victoria being the Kerang tragedy in June 2007. In Victoria, fines
increased for motorists that speed to ‘beat a train’, cross when lights and bells are
activated, or weave between boom gates, to $3304 (30 penalty units), 4 demerit points
and an automatic 3 month licence suspension. Such legislative initiatives together
with increased police enforcement will hopefully provide some level of deterrence to
motorists. However, before moving ahead with either educational or enforcement
initiatives, an analysis of the benefits and the costs of implementing any educational
or awareness raising campaign should be undertaken. Assessing the costs and
benefits may also serve as a basis for prioritising separate measures or measure
packages (such as the combination of education and enforcement) (SWOV Institute
for Road Safety Research, 2005).
308
REFERENCES
Aarts, H., & Dijksterhuis, A. (2000). Habits as knowledge structures: Automaticity in
goal-directed behavior. Journal of Personality and Social Psychology, 78, 53–
63.
Aarts, H., Verplanken, B., & van Knippenberg, A. (1998). Predicting behavior from
actions in the past: Repeated decision making or a matter of habit? Journal of
Applied Social Psychology, 28, 1355-1374.
Aberg, L., & Rimmo, P. (1998). Dimensions of aberrant driver behaviour.
Ergonomics, 41, 39-56.
Abraham, J., Datta, T. K., & Datta, S. (1998). Driver behaviour at rail-highway
crossings. Transportation Research Record, 1648, 28-34.
Adler, M., & Ziglio, E. (1996). Gazing into the oracle: The Delphi method and its
application to social policy and public health. London: Jessica Kingsley
Publishers.
Affleck Consulting Pty Ltd. (2003). The Australian rail industry: Overview and
issues. Melbourne: National Road Transport Commission.
Afxentis, D. (1994). Urban railway level crossings: Civil engineering working paper.
Melbourne: Monash University.
Ajzen, I. (1985). From intention to actions: A theory of planned behaviour. In J. K. J.
Beckman (Ed.), Action control: From cognitions to behaviour (pp. 11-39).
New York: Springer-Verlag.
Ajzen, I. (1991). The theory of planned behaviour. Organizational Behaviour and
Human Decision Processes, 50, 179-211.
Ajzen, I. (2002a). Perceived behavioral control, self-efficacy, locus of control, and the
theory of planned behavior. Journal of Applied Social Psychology, 32, 1-20.
309
Ajzen, I. (2002b). Residual effects of past on later behavior: Habituation and reasoned
action perspectives. Personality and Social Psychology Review, 6(2), 107–
122.
Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social
behavior. Englewood Cliffs, NJ: Prentice Hall.
Ajzen, I., & Fishbein, M. (2000). Attitudes and the attitude-behavior relation:
Reasoned and automatic processes. In W. Stroebe, & M. Hewstone, (Ed.),
European review of social psychology (pp. 1-33). Chichester, England: Wiley.
Ajzen, I., & Madden, T. J. (1986). Prediction of goal-directed behaviour: Attitudes,
intentions and perceived behavioural control. Journal of Experimental Social
Psychology, 22, 454-474.
Akers, R. L. (1990). Rational choice, deterrence and social learning theory in
criminology: The path not taken. The Journal of Criminal Law and
Criminology, 81(3), 653-676.
Akers, R. L. (1994). Criminological theories: Introduction and evaluation. Los
Angeles: Roxbury Publishing Company.
Akers, R. L., & Lee, G. (1996). A longitudinal test of social learning theory:
Adolescent smoking. Journal of Drug Issues, 26(2), 317-342.
Akers, R. L., Krohn, M. D., Lanza-Kaduce, L., & Radosevich, M. (1979). Social
learning and deviant behavior: A specific test of a general theory. American
Sociological Review, 44, 636-655.
Akers, R. L., Krohn, M. D., Lanza-Kaduce, L., & Radosevich, M. (1979). Social
learning and deviant behavior: A specific test of a general theory. American
Sociological Review, 44, 636-655.
310
Akers, R. L., & Lee, G. (1996). A longitudinal test of social learning theory:
Adolescent smoking. Journal of Drug Issues, 26(2), 317-342.
Alexander, G. J. (1989). Search and perception-reaction time at intersections and
railroad grade crossings. ITE Journal, 59(11), 17-20.
American Medical Association. (1998). A tale of two stories: Contrasting views of
patient safety. Report from a workshop on assembling the scientific basis for
progress on patient safety. Chicago: National Patient Safety Foundation at the
AMA.
Andrea, D. J., Fildes, B. N., & Triggs, T. J. (2001). The sensitivity and bias of older
driver judgements in an arrival-time task. Paper presented at the Road Safety
Research, Policing and Education Conference, Melbourne.
Armsby, P., Boyle, A. J., & Wright, C. C. (1989). Methods for assessing drivers’
perception of specific hazards on the road. Accident Analysis & Prevention,
21, 45-60.
Arnett, J. (2002). Developmental sources of crash risk in young drivers. Injury
Prevention, 8, 17-23.
ARRB Transport Research. (2002). Reducing collisions at passive railway level
crossings in Australia (No. RC2160-2). Sydney: AustRoads.
Athanaselis, S., Dona, A., Papadodima, S., Papoutsis, G., Maravelias, C., &
Koutselinis, A. (1999). The use of alcohol and other psychoactive substances
by victims of traffic accidents in Greece. Forensic Science International, 102,
103–109.
Attebo, K., Mitchell, P., & Smith, W. (1996). Visual acuity and the causes of visual
loss in Australia. The Blue Mountains Eye Study. Ophthalmology, 103, 357-
364.
311
Aurelius, J. P., & Korobow, N. (1971). The visibility and audibility of trains
approaching rail-highway grade crossings: (No. FRA-RP-71-1). Washington,
DC.: Federal Railroad Administration, US Department of Transportation.
Australasian Railways Association. (2006). National railway level crossing
behavioural plan (November/05). Canberra: ARA.
Australian Rail Association Incorporated. (2002). Rail facts sheet. No. 1: The rail
industry in Australia. Melbourne: Australian Rail Association Incorporated.
Australian Rail Association Incorporated. (2004a). Policies: New directions for rail.
from http://www.ara.net.au/policy/directions.php
Australian Rail Association Incorporated. (2004b). Rail industry economic profile.
Retrieved 1st May, 2004, from http://www.ara.net.au/society/economics.php
Australian Transport Council. (2000). National road safety action plan 2003 and
2004. Canberra: Australian Transport Council.
Australian Transport Council. (2002). Independent review of rail safety arrangements
in Australia. from http://www.atcouncil.gov.au/rail/
Australian Transport Council. (2003). National railway level crossing safety strategy.
Canberra: Australian Transport Council.
Australian Transport Safety Bureau. (2002a). Collision between the passenger train
5AL8 and vehicles at the Salisbury Interchange level crossing, Salisbury,
South Australia. Canberra: Australian Transport Safety Bureau.
Australian Transport Safety Bureau. (2002b). Level crossing accidents: Monograph
10. Canberra: Commonwealth Department of Transport and Regional
Services.
Australian Transport Safety Bureau. (2003). Level crossing accident fatalities.
Canberra: Australian Transport Safety Bureau.
312
Australian Transport Safety Bureau. (2004). Rail occurences in Australia. Canberra:
Australian Transport Safety Bureau.
Avery, G. C. (1973). The role of communications and propaganda in traffic safety
(No. 3/73). Sydney: Traffic Accident Research Unit: Department of Motor
Transport.
Bagozzi, R. P. (1981). Attitudes, intentions, and behavior: A test of some key
hypotheses. Journal of Personality and Social Psychology, 41, 607–627.
Bagozzi, R. P. (1992). The self-regulation of attitudes, intentions and behavior. Social
Psychology Quarterly, 55, 178-204.
Bailey, T. (2003). Differences in teaching approaches among car driving instructors.
Paper presented at the Road Safety Research, Policing and Education
Conference, Sydney.
Ball, K., & Owsley, C. (1991). Identifying correlates of accident involvement for the
older driver. Human Factors, 33(5), 583-595.
Ball, K., Owsley, C., Sloane, M., Roeneker, D., & Bruni, J. (1993). Visual attention
problems as a predictor of vehicle crashes in older drivers. Investigative
Opthalmology and Visual Science, 34, 3110-3123.
Bamberg, S., Ajzen, I., & Schmidt, P. (2003). Choice of travel mode in the theory of
planned behavior: The roles of past behavior, habit, and reasoned action.
Basic and Applied Social Psychology, 25(3), 175–187.
Bandura, A. (1977a). Self-efficacy: Toward a unifying theory of behavioral change.
Psychological Review, 84, 191-215.
Bandura, A. (1986). Social foundations of thought and action: a social cognitive
theory. Englewood Cliffs: Prentice Hall.
313
Bandura, A. (1989). Human agency in social cognitive theory. American
Psychologist, 44, 1175-1184.
Bandura, A. (1997). Self-efficacy: The exercise of control. New York: Freeman.
Bandura, A. (1998). Health promotion from the perspective of social cognitive theory.
Psychology and Health, 13, 623-649.
Bandura, A., Adams, N.E., & Beyer, J. (1977b). Cognitive processes mediating
behavioral change. Journal of Personality and Social Psychology, 35, 125-
139.
Barbour, R. S. (1999). The case for combining qualitative and quantitative approaches
in health services research. Journal of Health Services Research Policy, 4, 39–
43.
Bartholomew, K. L., Parcel, G. S., Kok, G., & Gottlieb, N. H. (2001). Intervention
mapping: Designing theory and evidence-based health promotion programs.
California: Mayfield Publishing Company.
Bartholomew, L. K., Parcel, G. S., & Kok, G. (1998). Intervention mapping: A
process for developing theory and evidence-based health education programs.
Health Education and Behaviour, 25(5), 545-563.
Bartholomew, L. K., Parcel, G. S., Kok, G., & Gottlieb, N. H. (2006). Planning health
promotion programs: An intervention mapping approach: Jossey-Bass.
Beirness, D. J., & Simpson, H. M. (1988). Lifestyle correlates of risky driving and
accident involvement among youth. Alcohol, Drugs and Driving, 4, 193-204.
Beirness, D. J., Simpson, H. M., & Desmond, K. (2003). The road safety monitor:
Highway/railway crossing safety. Ottawa: Traffic Injury Research Foundation.
Bell, N. S., Amoroso, P.J., Yore, M.M., Smith, G.S., & Jones, B.H. (2000). Self-
reported risk-taking behaviours and hospitalisation for motor vehicle injury
314
among active duty army personnel. American Journal of Preventative
Medicine, 18(3), 85–95.
Benekohal, R., Michaels, R., Shim, E., & Resende, P. (1994). Effects of aging on
older drivers’ travel characteristics. Transportation Research Record, 1438,
91-98.
Bentler, P. D., & Speckart, G. (1979). Models of attitude-behavior relations.
Psychological Review, 86(5), 452-464.
Bentler, P. D., & Speckart, G. (1981). Attitudes ‘cause’ behaviors: a structural
equation analysis. Journal of Personality & Social Psychology, 40, 226-238.
Berg, W. D., Fuchs, C., & Coleman, J. (1981). Evaluating the safety benefits of
railroad advance
warning signs. Transportation Research Record, 773, 1-6.
Berg, W. D., Knoblauch, K., & Hucke, W. (1982). Causal factors in railroad-highway
grade crossing accidents. Transportation Research Record, 847, 47-54.
Beyea, S., & Nicoll, L. (2000). Learn more using focus groups. Association of
Operating Room Nurses Journal, 71(4), 897-900.
Blincoe, K., Jones, A. P., Sauerzapf, V., & Haynes, R. (2006). Speeding drivers'
attitudes and perceptions of speed cameras in rural England. Accident Analysis
and Prevention, 38(2), 371-378.
Blumer, H. (1955). Attitudes and the social act. Social Problems, 3, 59-65.
Brainin, P. (1980). Safety and mobility issues in licensing and education of older
drivers. (No. Report DOT HS-805 492. NHTSA.). Washington DC: US
Department of Transport.
315
Brown, I. D. (1991). Highway hypnosis: Implications for road safety researchers and
practitioners. In A. Gale (Ed.), Vision in Vehicles III (pp. 459-466).
Amsterdam: Elsevier.
Brown, I. D., & Groeger, J.A. (1988). Risk perception and decision making during the
transition between novice and experienced driver status. Ergonomics, 31, 585–
597.
Brown, J., Horberry, T., Anderson, J., Regan, M., & Triggs, T. (2003). Investigation
of the effects of driver distraction. Paper presented at the Road Safety
Research, Policing and Education Conference, Sydney.
Bryman, A. (1992). Quantitative and qualitative research: further reflections on their
integration. In J. Brannen (Ed.), Mixing Methods: Qualitative and Quantitative
Research (pp. 57–78). Avebury: Aldershot.
Bureau of Transport and Regional Economics. (2002). Rail accident costs in
Australia. Canberra: Bureau of Transport and Regional Economics.
Bureau of Transport and Regional Economics. (2002). Rail accident costs in
Australia. Canberra: Bureau of Transport and Regional Economics.
Bureau of Transport Economics. (2000). Road crash costs in Australia (No. 102).
Canberra: Bureau of Transport Economics.
Burgess-Limerick, R., & Bowen-Rotsaert, D. (2002). Fatigue Management Program:
Pilot Evaluation (No. Phase 2 Wave 3 Report): Global Institute of Learning
and Development Consortium.
Burns, P. C., & Wilde, G. J. S. (1995). Risk taking in male taxi drivers: Relationships
among personality, observational data and driver records. Personality and
Individual Differences, 18, 267–278.
316
Buzeman, D. G., Viano, D. C., & Lovsund, P. (1998). Car occupant safety in frontal
crashes: A parameter study of vehicle mass, impact speed and inherent
vehicle protection. Accident Analysis and Prevention, 30(6), 713-722.
Caird, J. K., Creaser, J.I., Edwards, C.J., & Dewar, R.E. (2002). A human factors
analysis of highway-railway grade crossing accidents in Canada. Calgary:
Transportation Development Centre, Transport Canada.
Cairney, P. (2003). Prospects of improving the conspicuity of trains at passive
railway crossings (No. CR 217). Canberra: Australian Transport Safety
Bureau.
Cairney, P., Cornwell, D., & Mabbott, N. (2002). Conspicuity of enhanced lighting
treatments for railway locomotives (No. RC07278). Melbourne: ARRB
Transport Research.
Cairney, P., Gunatillake, T., & Wigglesworth, E. (2002). Reducing collisions at
passive railway crossings in Australia (No. Ap-R208.02.). Sydney: Austroads.
Cameron, M., & Newstead, S. (2000). Response by MUARC to Reinvestigation of the
effectiveness of the Victorian TAC road safety campaign. Melbourne: Monash
University Accident Research Centre.
Cameron, M., Cavallo, A., & Gilbert, A. (1992). Evaluation of the speed camera
program in Victoria 1990-91, Phase 1: General effects. Phase 2: Effects of
program mechanisms. (No. 42). Melbourne: Monash University Accident
Research Centre.
Cameron, M., Haworth, N., Oxley, J., Newstead, S., & Le, T. (1993). Evaluation of
Transport Accident Commission road safety television advertising.
Melbourne: Monash University Accident Research Centre.
317
Campbell, D. T. (1963). Social attitudes and other acquired behavioral dispositions. In
S. Koch (Ed.), Psychology: A study of a science (Vol. 6, pp. 94-172). New
York: McGraw-Hill.
Carberry, T., Wood, J., & Watson, B. (2004). Eye disease and driving performance:
Correlates between insight and capability in drivers with glaucoma. Paper
presented at the Road Safety Research, Policing and Education Conference.
Carp, F. M. (1988). Significance of mobility for the well-being of the elderly. In T. R.
Board (Ed.), Transportation in an aging society: Improving mobility and
safety for older persons. Washington, DC: National Research Council.
Carroll, A., Multer, J., & Markos, S. (1995). Safety of highway-railroad grade
crossings: Use of auxiliary external alerting devices to improve locomotive
conspicuity. Cambridge: U.S. Department of Transportation.
Cater, K., Chalmers, A., & Ledda, P. (2002). Selective quality rendering by exploiting
human inattentional blindness: looking but not seeing. Paper presented at the
ACM symposium on Virtual reality software and technology, Hong Kong.
Cavallo, A., & Triggs, T. (1998). Young driver research strategy. Melbourne
Australia: Monash University Accident Research Centre.
Cavallo, A., & Cameron, M. (1992). Evaluation of a random breath testing initiative
in Victoria 1990 and 1991: Summary report (No. 39). Melbourne: Monash
University Accident Research Centre.
Charlton, J., Oxley, J., Fildes, B., Oxley, P., Newstead, S., O’Hare, M., et al. (2003).
An investigation of self-regulatory behaviours of older drivers (No. 208).
Melbourne: Monash University Accident Research Centre.
318
Charlton, J., Oxley, J., Fildes, B., & Les, M. (2001a). Self-regulatory behaviour of
older drivers. Paper presented at the Road Safety Research, Policing and
Education Conference, Melbourne.
Charlton, J., Oxley, J., Fildes, B., & Les, M. (2001b). Self regulatory behaviour of
older drivers. Paper presented at the Road Safety Research, Policing and
Education Conference, Melbourne.
Christie, R. (2001). The effectiveness of driver training as a road safety measure: A
review of the literature (No. 01/03). Melbourne: Royal Automobile Club of
Victoria.
Christie, R. (2002). Road safety education and training from a public health
perspective. Paper presented at the Road Safety Research, Policing and
Education Conference, Sydney.
CityRail. (2005). Cameras at level crossings. Retrieved 29th September, from
http://www.cityrail.info/security/cctv_level_crossings.jsp
Coghlan, M. (1997). Grade crossing safety issues: TDG’s Bus acceleration study and
truck acceleration study: Government of Canada.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences. Hillsdale, NJ:
Lawrence Erlbaum Associates.
Colman, A. (1995). Psychological research methods and statistics. New York:
Longman Group.
Commonwealth of Australia. (2002). Code of practice for the defined interstate rail
network. Volume 5: Rollingstock (Part 5: Specific requirements for
locomotives). Canberra: Australian Rail Operations Unit (Australian
Government Department of Transport and Regional Services).
319
Conner, M., & Sparks, P. (1996). The theory of planned behaviour and health
behaviours. In M. Conner, & Norman, P. (Ed.), Predicting health behaviour
(pp. 121– 162). Buckingham: Open University Press.
Cook, T., & Campbell, D. (1979). Quasi-experimentation: Design and analysis issues
for field settings. Chicago: Rand McNally.
Corbett, C. (1995). Road traffic offending and the introduction of speed cameras in
England: The first self-report study. Accident Analysis & Prevention, 27(3),
345–354.
Corbett, C., & Simon, F. (1999). The effects of speed cameras: How drivers respond
(No. 11). London, England: DTLR.
Crutcher, J., Black, G., & Campbell, P. (1994). Risky driving behaviours among
teenagers: Gwinnett County, Georgia, 1993. Journal of American Medical
Association, 272, 844–845.
Custer, R. L., Scarcella, J. A., & Stewart, B. R. (1999). The modified delphi technique
- A rotational modification. Journal of Vocational and Technical Education,
15(2).
Daigneault, G., Joly, P., & Frigon, J. (2002). Previous convictions or accidents and
the risk of subsequent accidents of older drivers. Accident Analysis and
Prevention, 34, 257–261.
Darzentas, J., & McDowell, M. R. C. (1981). Driver behaviour at unlit non-urban T-
junctions in daylight and darkness. Journal of the Operations Research
Society of America, 32, 721-727.
Davey, J., Wishart, D., Freeman, J., & Watson, B. (2007). An application of the
Driver Behaviour Questionnaire in an Australian organisational fleet setting.
320
Transportation Research Part F: Traffic Psychology and Behaviour, 10(1),
11-21.
Deery, H. A. (1999). Hazard and risk perception among young novice drivers.
Journal of Safety Research, 30(4), 225–236.
Deery, H. A., & Love, A.W. (1996). The effect of a moderate dose of alcohol on the
traffic hazard perception profile of young drivers. Addiction,, 91, 815–827.
Degenhardt, L., Roxburgh, A., Black, E., & Dunn, M. (2006). 2004 Cocaine and
Amphetamine Related Drug-Induced Deaths in Australia. Sydney: National
Drug and Alcohol Research Centre.
Dejoy, D. M. (1992). An examination of gender differences in traffic accident
perception. Accident Analysis & Prevention, 21, 155–168.
Dekker, S. (2002). Reconstructing human contributions to accidents: The new view
on error and performance. Journal of Safety Research, 33, 371-385.
Del Rio, C. M., Gomez, J., Sancho, M., & Alvarez, F. J. (2002). Alcohol, illicit drugs
and medicinal drugs in fatally injured drivers in Spain between 1991 and
2000. Forensic Science International, 127, 63–70.
Delaney, A., Lough, B., Whelan, M., & Cameron, M. (2004). A review of mass media
campaigns in road safety. Melbourne: Monash University Accident Research
Centre.
Department for Transport Energy and Infrastructure. (2006). Railway level crossing
safety education campaign. Retrieved 3rd January 2006, from
www.transport.sa.gov.au/safety/rail/advertising_campaign.asp
Department of Primary Industries and Energy. (1994). Canberra: Commonwealth of
Australia.
321
Department of Transport and Regional Services. (2004). Rail policy and programs.
from http://www.dotars.gov.au/transreg/str_about.htm
deVries, H., Kok, G., & Dijkstra, M. (1990). Self-efficacy as a determinant of the
onset of smoking and interventions to prevent smoking in adolescents. In R. J.
Takens (Ed.), European Perspectives in Psychology (Vol. 2). London: Wiley
& Sons.
Dewer, R. E. (2002). Railroad grade crossing accidents. In R. E. Dewer, & Olson,
P.E., (Ed.), Human factors in traffic safety (pp. 507-523). Tucson, Arizona:
Lawyers & Judges Publishing Company Inc.
Di Pietro, G., & Ivett, L. (2003). Road Safety Education and Training from a Public
Health Perspective. Paper presented at the Road Safety Research, Policing and
Education Conference, Sydney.
Di Stefano, M., & Macdonald, W. (2003). Assessment of older drivers: Relationships
among on-road errors, medical conditions and test outcome. Journal of Safety
Research, 34, 415– 429.
Diamantopoulou, K., Skalova, M., Dyte, D., & Cameron, M. (1996). Crash risks of
road user groups in Victoria (No. 88). Melbourne: Monash University
Accident Research Centre.
Dobson, A., Brown, W., Ball, J., McFadden, M., & Walker, M. (1998). Female driver
behaviour and road crash involvement. Paper presented at the Conference on
Road Safety, Perth.
Doll, J., & Ajzen, I. (1992). Accessibility and stability of predictors in the theory of
planned behaviour. Journal of Personality and Social Psychology, 63(5), 754-
765.
322
Donohue, S. E., Wendelken, C., Crone, E. A., & Bunge, S. A. (2005). Retrieving rules
for behavior from long-term memory. NeuroImage, 26(4), 1140 – 1149.
Donovan, R. J., Jalleh, G., & Henley, N. (1999). Executing effective road safety
advertising: Are big production budgets necessary? Accident Analysis &
Prevention, 31(3), 243-252.
Drakopoulos, A., & Lyles, R. (2001). An evaluation of age effects on driver
comprehension of flashing traffic signal indications using multivariate
multiple response analysis of variance models. Journal of Safety Research, 32,
85-106.
Drever, E. (1995). Using semi-structured interviews in small-scale research. A
teacher's guide. Edinburgh: Scottish Council for Research in Education.
Drummer, O. (2002). Fatal driver study 1990-1999: Report to VicRoads. Melbourne:
Victorian Institute of Forensic Medicine.
Drummer, O., Gerostamoulous, J., Batziris, H., Chu, M., Caplehorn, J., Robertson,
M.D., &, & Swann, P. (2004). The involvement of drugs in drivers of motor
vehicles killed in Australian road traffic crashes. Accident Analysis and
Prevention, 36, 239-248.
Drummond, A. E., & Yeo, E. (1992). The risk of driver crash involvement as a
function of driver age. (No. 49). Melbourne, Australia: Monash University
Accident Research Centre.
Duperrex, O., Bunn, F., & Roberts, I. (2002). Safety education of pedestrians for
injury prevention: a systematic review of randomised controlled trials. British
Medical Journal, 324, 1129-1129.
323
Durkin, M., Laraque, D., Lubman, I., & Barlow, B. (1999). Epidemiology and
prevention of traffic injuries to urban children and adolescents. Paediatrics,
103, e74.
Eberhard, J. (1996). Safe mobility of senior citizens. Journal of International
Association of Traffic and Safety Sciences, 20(1), 29-37.
Eby, D. W., & Kostyniuk, L. P. (1998a). Maintaining older driver mobility and well-
being with traveler information systems. Transportation Quarterly, 52(4), 45–
53.
Eby, D. W., Trombley, D.A., Molnar, L.J., & Shope, J.T. (1998b). The assessment of
older drivers’ capabilities: A review of the literature. Ann Arbor, MI: The
University of Michigan Transportation Research Institute.
Economic Associates Pty Ltd. (2002). NHVAS Fatigue management module: Draft
regulatory impact statement. Brisbane: Queensland Transport.
Ehrlich, D. (1989). Driver behaviour and decision making. Paper presented at the
National Conference on Rail-Highway Safety, San Diego.
Elander, J., West, R., & French, D. (1993). Behavioral correlates of individual
differences in road-traffic crash risk: An examination of methods and findings.
Psychological Bulletin, 113, 279–294.
Elkington, J., & Hunter, K. (2003). Expanding our Concept of Best Practice in Road
Safety Education: A Review of Current Evidence and Practice. Paper
presented at the Road Safety Research, Policing and Education Conference,
Sydney.
Elliott, B. (1992). Report on achieving high levels of compliance with road safety
laws: A review of road user behaviour modification (No. 6). Brisbane: Travel
Safe Committee, Legislative Assembly of Queensland.
324
Elliott, B. (2003a). Deterrence theory revisted. Paper presented at the Road Safety
Research, Policing and Education Conference, Sydney, N.S.W.
Elliott, M. A., Armitage, C.J., & Baughan, C.J. (2003b). Drivers' compliance with
speed limits: An application of the Theory of Planned Behavior. Journal of
Applied Psychology, 88(5), 964-972.
Elvik, R. (2004a). To what extent can theory account for the findings of road safety
evaluation studies? Accident Analysis & Prevention, 36, 841–849.
Elvik, R., Christensen, P., & Amundsen, A. (2004b). Speed and road accidents. An
evaluation of the Power Model. (No. 740/2004). Oslo, Sweden: Institute of
Transport Economics TOI.
Erickson, M. L., Gibbs, J.P. & Jensen, G.F. (1977). The deterrence doctrine and the
perceived certainty of legal punishments. American Sociological Review, 42,
305-317.
Evans, D., & Norman, P. (1998). Understanding pedestrians' road crossing decisions:
An application of the theory of planned behaviour. Health Education
Research, 13, 481-489.
Evans, D., & Norman, P. (2003). Predicting adolescent pedestrians' road- crossing
intentions: An application and extension of the Theory of Planned Behaviour.
Health Education Research, 18, 267-277.
Eysenck, H. J. (1983). A biometrical-genetical analysis of impulsive and sensation
seeking behaviour. In M. Zuckerman (Ed.), Biological Bases of Sensation
Seeking, Impulsivity, and Anxiety. Hillside, NJ: Lawrence Erlbaum Associates.
Fatigue Expert Group. (2001). Options for regulatory approach to fatigue in drivers
of heavy vehicles in Australia and New Zealand: Prepared for the National
325
Road Transport Commission, the Australian Transport Safety Bureau and the
New Zealand Land Transport Authority,.
Fattah, E. (1976). Deterrence: A review of the literature: Law Reform Commission of
Canada.
Fazio, R. H. (1990). Multiple processes by which attitudes guide behavior: The
MODE model as an integrative framework. In M. P. Zanna (Ed.), Advances in
experimental social psychology (pp. 75–109). San Diego, CA: Academic.
Federal Office of Road Safety. (1996). Female car drivers and risk. Monograph 12.
Canberra: FORS.
Federal Office of Road Safety. (1997). Road fatalities Australia: 1996 statistical
summary. Canberra: FORS.
Federal Railroad Administration. (2001). Railroad safety statistics: Annual report.
Washington, D.C.: Federal Railroad Administration.
Federal Railroad Administration. (2004). Grade crossings. Retrieved 6th July, 2004,
from http://www.fra.dot.gov/Content3.asp?P=227
Ferguson, M., Sheehan, M., Davey, J., & Watson, B. (1999). Drink driving
rehabilitation: The present context. Report prepared for the Australian
Transport Safety Bureau. (No. CR 184). Brisbane: CARRS-Q, Queensland
University of Technology.
Fergusson, D., Swain-Campbell, N., & Horwood, J. (2003). Risky driving behaviour
in young people: Prevalence, personal characteristics and traffic accidents.
Australian and New Zealand Journal of Public Health, 27(3), 337-342.
Feyer, A.-M., Williamson, A. M., & Friswell, R. (1995). Strategies to combat driver
fatigue in the long distance road transport industry. Stage 2: Evaluation of
two-up operations. (No. CR 158). Canberra: Federal Office of Road Transport.
326
Fieldwick, R. (1993). The relationship between rural speed limit and accident rate. In
B. Fildes & S. Lee (Eds.), The speed review: Road environment, behaviour,
speed limits, enforcement and crashes. Melbourne, Australia: Monash
University Accident Research Centre.
Fildes, B. (1997). Safety of older drivers: Strategy for future research and action
initiatives. Melbourne: Monash University Accident Research Centre.
Fildes, B. (2004). Overview of older user safety statistics and research. In Centre for
Accident Research and Road Safety - Queensland (Queensland University of
Technology) (Ed.), Road Safety Issues for Older Road Users: Monograph 1
(pp. 11-20). Brisbane.
Fildes, B., Corben, B., Kent, S., Oxley, J., Le, T., & Ryan, P. (1994). Older road user
crashes: Driver, pedestrian, elderly, injury, crash trends, crash
characteristics, road trauma, costs, countermeasures (No. 61). Melbourne:
Monash University Accident Research Centre.
Fildes, B., Corben, B., Morris, A., Oxley, J., Pronk, N., Brown, L., et al. (2000). Road
safety environment and design for older drivers (No. AP-169/00). Sydney:
Austroads.
Fildes, B., Fitzharris, M., Charlton, J., & Pronk, N. (2001). Older driver safety – a
challenge for Sweden’s ‘Vision Zero’. Paper presented at the Australian
Transport Research Forum, Hobart.
Finn, P., & Bragg, B. W. E. (1986). Perception of the risk of an accident by young
and older drivers. Accident Analysis & Prevention, 18(4), 289–298.
Fishbein, M. (1967). Readings in attitude theory and measurement. New York: Wiley.
Fishbein, M. (2000). The role of theory in HIV prevention. AIDS Care, 12, 273-278.
327
Fishbein, M. (2001). Project SAFER: using theory to identify critical targets for HIV
prevention interventions. Psychology, Health and Medicine, 6(2), 137-138.
Fishbein, M. (2003). Models of health behaviour. Paper presented at the Behavioural
Approaches to Injury Control, Seattle, Washington.
Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An
introduction to theory and research. Reading, MA: Addison-Wesley.
Fishbein, M., Hennessy, M., Kamb, M., Bolan, G., Hoxworth, T., Iatesta, M., et al.
(2001). Using intervention theory to model factors influencing behavior
change: Project RESPECT. Evaluation and the Health Professions, 24, 363-
384.
Fishbein, M., Hennessy, M., Yzer, M., & Douglas, J. (2003). Can we explain why
some people do and some people do not act on their intentions? Psychology,
Health and Medicine, 8(1), 3-18.
Fitzpatrick, K., Bartoskewitz, R., Bean, J., & Carlson, P. (1997). Traffic violations at
gated highway-railroad grade crossings. College Station, Texas: Texas
Transportation Institute.
Ford, G., & Matthews, A. (2002). Analysis of Australian grade crossing accident
statistics. Paper presented at the 7th International Symposium on Rail-road
Highway Grade Crossing Research and Safety, Melbourne, Monash
University.
Foreman, E. K. (1991). Survey sampling principles. New York: M. Dekker.
Freund, B., Colgrove, L. A., Burke, B. L., & McLeod, R. (2005). Self-rated driving
performance among elderly drivers referred for driving evaluation. Accident
Analysis & Prevention, 37.
328
Frith, B., Strachan, G., & Patterson, T. (2005). Road safety implications of excessive
and inappropriate vehicle speed. In J. Langford & B. Fildes (Eds.), Austroads
research report. Australasian road safety handbook: Volume 2. Sydney.
Frith, W. (2003). An investigation of the relationship between speed enforcement,
vehicle speeds and injury crashes in New Zealand. Paper presented at the
Road Safety Research, Policing and Education Conference.
Frith, W. J., & Toomath, J. B. (1982). The New Zealand open road speed limit.
Accident Analysis and Prevention, 14(3), 209-218.
Gardner, M., & Steinberg, L. (2005). Peer influence on risk taking, risk preference,
and risky decision making in adolescence and adulthood: An experimental
study. Developmental Psychology, 41(4), 625-635.
Gärling, T. (1992). Determinants of everyday time allocation. Scandinavian Journal
of Psychology, 33, 160-169.
George, R. M. (2002a). Austroads heavy vehicle safety projects. Melbourne: ARRB
Transport Research Ltd.
George, R. M. (2002b). Vehicle compliance with speed and mass: Evidence from
weigh-in-motion devices. Report to Austroads. Melbourne: ARRB Transport
Research Ltd.
Giambra, L. M., Camp, C. J., & Grodsky, A. (1992). Curiosity and stimulation
seeking across the adult life span: Cross-sectional and seven-year longitudinal
findings. Psychology and Aging, 7, 150–157.
Gillholm, R., Erdeus, J., & Garling, T. (2000). The effect of choice on intention-
behavior consistency. Scandinavian Journal of Psychology, 41(1), 1-8.
Gillholm, R., Ettema, D., Selart, M., & Gärling, T. (1996). The role of planning for
intention-behavior consistency. Göteborg Psychological Reports, 26(8), 1-20.
329
Glendon, A. I., Dorn, L., Davies, D. R., Matthews, G., & Taylor, R. G. (1996). Age
and gender differences in perceived accident likelihood and driver
competences. Risk Analysis, 16(6), 755–762.
Glennon, J. C., & Engr, D. (2005). Rail-highway grade crossing design and
maintenance. In J. Loumiet & W. Jungbauer (Eds.), Train accident
reconstruction and FELA and railroad litigation. Tucson, Arizona: Lawyers
and Judges Publishing Company.
Global Road Safety Partnership. (2006). Safer road users. Retrieved 17th February,
from http://www.grsproadsafety.org/?pageid=110
Goldenbeld, C., Levelt, P., & Heidstra, J. (2000). Psychological perspectives on
changing driver attitude and behaviour. Recherche Transports Securite, 67,
65-81.
Goldenbeld, C., & van Schagen, I. (2005). The effects of speed enforcement with
mobile radar on speed and accidents An evaluation study on rural roads in the
Dutch province Friesland. Accident Analysis and Prevention, 37(6), 1135–
1144.
Gollwitzer, P. M. (1990). Action phases and mindsets. In E. T. Higgins & J. R. M.
Sorrentino (Eds.), The handbook of motivation and cognition (Vol. 2, pp. 53-
92). New York: Guilford.
Gordon, C., & Hunt, M. (1998). The theory of planned behaviour applied to speeding,
drink-driving and seat belt wearing. Paper presented at the Road Safety
Research, Policing and Education Conference, Wellington, New Zealand.
Gossner, O., & Picard, P. (2005). On the consequences of behavioural adaptations in
the cost-benefit analaysis of road safety measures. The Journal of Risk and
Insurance, 72(4), 577-599.
330
Gou, M., & Bellvigna-Ladoux, O. (2003). Impact of heavy vehicles on crossing
safety: Development of an adapted design tool. Montreal: Transport Canada.
Green, M. (2002). Signs and signals. Occupational Health and Safety, 71, 30-36.
Gregersen, N. P., & Bjurulf, P. (1996). Young novice drivers own skill: Towards a
model of their accident involvement. Accident Analysis & Prevention, 28(2),
229-241.
Groeger, J. A., & Brown, I. D. (1989). Assessing one’s own and others’ driving
ability: The influences of sex, age and experience. Accident Analysis &
Prevention, 21(2), 155–168.
Grossman, D., & Garcia, C. (1999). Effectiveness of health promotion programs to
increase motor vehicle occupant restraint use among young children.
American Journal of Preventive Medicine, 16, 12-22.
Guria, J. (1999). An economic evaluation of incremental resources to road safety
programmes in New Zealand. Accident Analysis and Prevention, 31, 91-99.
Gusfield, J. R. (1991). Risky roads. Society, 28, 10–16.
Haegerstrom-Portnoy, G., Schneck, M. E., & Brabyn, J. A. (1999). Seeing into old
age: Vision function beyond acuity. Optometry and Visual Science, 76, 141-
158.
Haid, K. (2002). Opportunities and limitations in road safety information: Swedish
National Road Administration.
Hall, G. (2002). Introducing Operation Lifesaver. Paper presented at the 7th
International Symposium on Railroad-Highway Grade Crossing Research and
Safety., Melbourne.
Hall, G. (2004). Grade crossing collision investigation (GCCI) courses: Bringing
local law enforcement into the rail safety solution. Paper presented at the 8th
331
International Level Crossing Symposium and Managing Trespass Seminar,
Sheffield, UK.
Hall, J., & West, R. (1996). The role of formal instruction and informal practice in
learning to drive. Ergonomics, 39, 693–706.
Hamelin, P. (1987). Lorry drivers' time habits in work and their involvement in traffic
accidents. Ergonomics, 30, 1323-1333.
Hanowski, R., Hickman, J., Fumero, M., Olson, R., & Dingus, T. (2007). The sleep of
commercial vehicle drivers under the 2003 hours-of-service regulations.
Accident Analysis & Prevention, In Press Corrected Proof.
Hardy, A. (2004). Audibility of warning horns. London: Rail Safety and Standards
Board.
Harre, N. (2000). Risk evaluation, driving, and adolescents: A typology.
Developmental Review, 20(2), 206–226.
Harrison, W. A. (1998). Applying psychology to a reluctant road safety: A comment
on South. Australian Psychologist, 33(3), 238-240.
Hartley, L. R., & Arnold, P. K. (1996). Management of fatigue in the road transport
industry. Paper presented at the 2nd International Conference on Fatigue in
Transportation,, Fremantle, WA.
Harwood, D. W. (1993). Use of rumble strips to enhance safety (No. 191).
Washington: Transport Research Board for AASHTO.
Hatakka, M., Keskinen, E., Katila, A., & Laapotti, S. (1997). Self-reported driving
habits are valid predictors of violations and accidents. In T. Rothengatter & E.
Carbonell Vaya (Eds.), Traffic & Transport Psychology: Theory and
Application (pp. 295-303). New York: Pergamon.
332
Haworth, N., Tingvall, C., & Kowadlo, N. (2000). Review of best practice road safety
initiatives in the corporate and/or business environment. Melbourne: Monash
University Accident Research Centre.
Haworth, N., Vulcan, P., & Sweatman, P. (2002). Truck safety benchmarking study:
Report prepared for the NRTC. Melbourne: National Road Transport
Commission.
Haworth, N. L., Triggs, T. J., & Grey, E. M. (1988). Driver fatigue: Concepts,
measurement and crash countermeasures (No. CR 72). Canberra: Federal
Office of Road Safety,.
Health and Safety Executive. (2005). Level crossings: Summary of findings and key
human factors issues (No. 359). Hertfordshire, England: Her Majesty's Rail
Inspectorate.
Henderson, M. (1991). Education, publicity and training in road safety: A literature
review (No. 22). Melbourne, Australia: Monash University Accident Research
Centre.
Hill, D. (2001). Issues affecting Australian rail. Paper presented at the Asia Pacific
Rail 2001, Singapore.
Hoelscher, D. M., Evans, A., Parcel, G. S., & Kelder, S. H. (2002). Designing
effective nutrition interventions for adolescents. Journal of the American
Dietetic Association, 102(Suppl 3), S52-63.
Hoffman, R. R., & Woods, D.D. (2000). Studying cognitive systems in context.
Human Factors, 42(1), 1 –7.
Holland, C., & Rabbitt, P. (1992). People’s awareness of their age-related sensory and
cognitive deficits and the implications for road safety. Applied Cognitive
Psychology, 6, 217-231.
333
Holland, C. A. (1993). Self-bias in older drivers’ judgments of accident likelihood.
Accident Analysis & Prevention, 25(4), 431–441.
Holland, C. A., & Conner, M. T. (1996). Exceeding the speed limit: An evaluation of
the effectiveness of a police intervention. Accident Analysis & Prevention,
28(5), 587–597.
Homel, R. (1986). Policing the drinking driver: random breath testing and the
process of deterrence. Canberra: Federal Office of Road Safety.
Homel, R. (1988). Policing and punishing the drinking driver: A study of specific and
general deterrence. New York.: Springer-Verlag.
Horswill, M. S., Waylen, A. E., & Tofield, M. I. (2004). Drivers' ratings of different
components of their own driving skill: a greater illusion of superiority for
skills that relate to accident involvement. Journal of Applied Social
Psychology, 34(1), 177-195.
Horvath, P., & Zuckerman, M. (1993). Sensation seeking, risk appraisal, and risky
behavior. Personality and Individual Differences, 14, 41–52.
House of Representatives: Standing Committee on Transport and Regional Services.
(2004). Train Illumination: Inquiry into some measures proposed to improve
train visibility and reduce level crossing accidents. Canberra: The Parliament
of the Commonwealth of Australia.
Howarth, C. I., & Repetto-Wright, R. (1978). Measurement of risk and the attribution
of responsibility for child pedestrian accidents. Safety Education, 144, 10-13.
Howarth, C. I., Routledge, D.A. & Repetto-Wright, R. (1974). An analysis of road
accidents involving young children. Ergonomics, 17, 319-330.
334
Hughes, P. (2002). A risk assessment system for passive level crossings. Paper
presented at the 7th International Symposium on Railroad-Highway Grade
Crossing Research and Safety, Monash University, Melbourne.
Hughes, P. K., & Coles, B. L. (1986). Can the conspicuity of objects be predicted
from laboratory experiments? Ergonomics, 29(9), 1097-1111.
Janke, M. (1994). Age-related disabilities that may impair driving and their
assessment: Literature review. Sacramento, United States: Department of
Motor Vehicles.
Janssen, W. H., & Tenkink, E. (1988). Considerations on speed selection and risk
homeostasis in driving. Accident Analysis and Prevention, 20(2), 137-142.
Jessor, R. (1977). Risky driving and adolescent problem behaviour. Alcohol, Drugs
and
Driving, 3, 1-12.
Jette, A. M., & Branch, L. G. (1992). A ten-year follow-up of driving patterns among
community dwelling elderly. Human Factors, 34, 25-31.
Johnston, I. (2004). Reducing injury from speed related road crashes. Injury
Prevention, 10(5), 257-259.
Jonah, B., & Dawson, N. (1987). Youth and risk: Age differences in risky driving,
risk perceptions and risk utility. Alcohol, Drugs and Driving, 3, 13-29.
Jonah, B. A. (1997). Sensation seeking and risky driving: A review and synthesis of
the literature. Accident Analysis and Prevention, 29(5), 651-665.
Jonah, B. A., Thiessen, R., & Au-Yeung, E. (2001). Sensation seeking, risky driving
and behavioural adaptation. Accident Analysis & Prevention, 33, 679-684.
Jones, J., & Hunter, D. (1995). Qualitative research: Consensus methods for medical
and health services research. British Medical Journal, 311, 376-380.
335
Jordon, P. (2006). A trial of a low cost level crossing warning device. Paper presented
at the IRSE Annual Meeting (17th February), Melbourne.
King, M. (2004). Road Safety Issues for Older Road Users: Monograph 1. Brisbane:
CARRS-Q, Queensland University of Technology.
Kitzinger, J. (1995). Introducing focus groups. British Medical Journal, 311, 299-302.
Kleinke, C. L. (1984). Two models for conceptualising the attitude – behavior
relationship. Human Relations, 37, 333-350.
Kline, D., Kline, T., Fozard, J., Kosnik, W., Scheiber, R., & Sekuler, R. (1992).
Vision, aging and driving: The problems of older drivers. Journal of
Gerontology: Psychological Sciences, 47(1), 27-34.
Knox, D. J., & Turner, B.M. (2002). Reaching a consensus: Using the Delphi method
to assess expert opinion on countermeasures against unlicensed driving. Paper
presented at the Behavioural Research in Road Safety: Twelfth Seminar,
London.
Koepsell, T. D., Wolf, M. E., McCloskey, L., Buchner, D. M., Louie, D., Wagner, E.
H., & Thompson, R. S. (1994). Medical conditions and motor vehicle collision
injuries in older adults. Journal of the American Geriatric Society, 42, 695–
700.
Kok, G., deVries, H., Mudde, A. N., & Strecher, V. J. (1991). Planned health
education and the role of self-efficacy: Dutch research. Health Education
Research, 6, 231-238.
Kok, G., Schaalma, H., Ruiter, R., Empelen, P., & Brug, J. (2004). Intervention
mapping: A protocol for applying Health Psychol theory to prevention
programmes. Journal of Health Psychology, 9, 85-89.
336
Kontogiannis, T., Kossiavelou, Z., & Marmaras, N. (2002). Self-reports of aberrant
behaviour on the roads: Errors and violations in a sample of Greek drivers.
Accident Analysis and Prevention, 24(3), 381-399.
Kraus, N. N., & Slovic, P. (1988). Taxonomic analysis of perceived risk: Modelling
individual and group perceptions. Risk Analysis, 8, 435-455.
Kraus, S. J. (1995). Attitudes and the prediction of behavior: A meta-analysis of the
empirical literature. Personality & Social Psychology Bulletin, 21, 58-75.
Krueger, G. P. (1989). Sustained work, fatigue, sleep loss and performance: A review
of issues. Work and Sleep, 3, 129-141.
Krueger, H.-P., & Vollrath, M. (2000). Effects of cannabis and amphethamines on
driving simulator performance of recreational drug users in the natural field.
Paper presented at the 15th International Conference on Alcohol, Drugs and
Traffic Safety (T2000), Stockholm, Sweden.
Laapotti, S., & Keskinen, E. (1998). Differences in fatal loss-of-control accidents
between young male and female drivers. Accident Analysis & Prevention,
30(4), 435-442.
Lajunen, T., & Rasanen, M. (2004a). Can social psycholocial models be used to
promote bicycle helmet use among teenagers? A comparison of the Health
Belief Model, Theory of Planned Behaviour and the Locus of Control. Journal
of Safety Research, 35, 115-123.
Lajunen, T., & Summala, H. (2003). Can we trust self-reports of driving? Effects of
impression management on driver behaviour questionnaire responses.
Transportation Research Part F: Traffic Psychology and Behaviour, 6(2), 97-
107.
337
Lajunen, T., Parker, D., & Summala, H. (2004b). The Manchester driver behaviour
questionnaire: A cross-cultural study. Accident Analysis & Prevention, 36(2),
231-238.
Lam, L. T. (2002). Distractions and the risk of car crash injury: The effect of drivers'
age. Journal of Safety Research, 33, 411-419.
Lammert, J. K. (1999). Application of automated enforcement for highway-railroad
grade crossings. Texas: Department of Civil Engineering, Texas A&M
University.
Lapinski, M. K., & Witte, K. (1998). Health communication campaigns. In L. D.
Jackson & B. K. Duffy (Eds.), Health communication research: A guide to
developments and directions. Westport, Conneticut: Greenwood Press.
Lawton, R., Parker, D., Manstead, A.S.R., & Stradling, S. G. (1997a). The role of
affect in predicting social behaviors: The case of road traffic violations.
Journal of Applied Social Psychology, 27, 1258– 1276.
Lawton, R., Parker, D., Stradling, S.G., & Manstead, A.S.R. (1997b). Predicting road
traffic accidents: The role of social deviance and violations. British Journal of
Psychology, 88(2), 249–263.
Lay, M. G. (1990). Handbook of road technology: Traffic and transport Volume 2
(2nd ed.). New York: Gordon and Breach Publishing Group.
Leadbeatter, C. (1997). RoadSmart: An evaluation., University of Melbourne, Masters
Thesis.
Lechner, L., & deVries, H. (1995). Starting participation in an employee fitness
program: Attitudes, social influence, and self-efficacy. Preventive Medicine,
24, 627-633.
338
Leibowitz, H. W. (1985). Grade crossing accidents and human factors engineering.
American Scientist, 73, 558.
Levy, D. T. (1990). Youth and traffic safety: The effects of driving age, experience
and education. Accident Analysis and Prevention, 22(4), 327-334.
Lex Service. (1997). Lex Report on Motoring. Driving for Safety. London, England:
Lex Service.
Linstone, H., & Turoff, M. (1975). The Delphi method: Techniques and applications.
Massachusetts: Addison-Wesley.
Lobb, B., Harre, N., & Suddendorf, T. (2001). An evaluation of a suburban railway
pedestrian crossing safety programme. Accident Analysis & Prevention, 33(2),
157-165.
Longo, M., Hunter, C., Lokan, R., White, J., & White, M. (2000). The prevalence of
alcohol, cannabinoids, benzodiazepines and stimulants amongst injured
drivers and their role in driver culpability. Part 1. The prevalence of
drug use in drivers, and characteristics of the drugpositive group. Accident, Analysis
and Prevention, 32, 613–622.
Lundberg, C., Hakamies-Blomqvist, L., Almkvist, O., & Johansson, K. (1998).
Impairments of some cognitive functions are common in crash-involved older
drivers. Accident Analysis and Prevention, 30(3), 371– 377.
Lyman, J. M., McGwin, G., & Sims, R. V. (2001). Factors related to driving difficulty
and habits in older drivers. Accident Analysis and Prevention,, 33, 413– 421.
Lyznicki, J. M., Doege, T. C., Davis, R. M., & Williams, M. A. (1998). Sleepiness,
driving, and motor vehicle crashes. Journal of the American Medical
Association, 279, 1908-1913.
339
Mabbott, N., Newman, S., & Moore, B. (2001). Safety and productivity through
flexibility: Driving hours review. Paper presented at the ARRB Transport
Research Ltd Conference, Melbourne.
MacIntyre, S., & Petticrew, M. (2000). Good intentions and received wisdom are not
enough. Journal of Epidemiology & Community Health, 54(11), 802-803.
Mack, A. (2003). Inattentional blindness: Looking without seeing. Current Directions
in Psychological Science, 12(5), 180-184.
Mack, A., & Rock, I. (1998). Inattentional blindness. Cambridge: MIT Press.
Masser, I., & Foley, P. (1987). Delphi revisited: Expert opinion in urban analysis.
Urban Studies, 24(3), 217-225.
Matthews, M. (1986). Aging and the perception of driving risk and ability. Paper
presented at the Human Factors Society, 30th Annual Meeting.
Maycock, G., Lockwood, C.R. & Lester, J.F. (1991). The accident liability of car
drivers. (No. RR315.). Crowthorne, United Kingdom: Transport Research
Laboratory.
Mayhew, D. R., & Simpson, H.M. (1995). The role of driving experience:
Implications for the training and licensing of new drivers. Toronto, Canada:
Insurance Bureau of Canada: Toronto.
Mayhew, D. R., & Simpson, H.M. (1996). Effectiveness and role of driver education
and training in a graduated licensing system. Ottawa: Traffic Injury Research
Foundation.
McBride, G., & Faulks, I. (2002, 7-9 August). Railway level crossing safety. Paper
presented at the Local Government Road Safety Conference, Newcastle.
McGraw, K. O., & Wong, S. P. (1996). Forming inferences about some intraclass
correlation coefficients. Psychological Methods, 1(1), 30-46.
340
McGwin, G., & Brown, D. B. (1999a). Characteristics of traffic crashes among
young, middle-aged and older drivers. Accident Analysis & Prevention, 31,
181-198.
McGwin, G., & Brown, D. B. (1999b). Characteristics of traffic crashes among
young, middle-aged and older drivers. Accident Analysis & Prevention, 31,
181-198.
McKenna, F. P. (1993). It won’t happen to me: Unrealistic optimism or illusion of
control? British Journal of Psychology, 84, 39-50.
Meadows, M., & Stradling, S. (1999). Are women better drivers than men? In J.
Hartley & A. Branthwaite (Eds.), The Applied Psychologist (2nd ed.).
Buckingham: Open University Press.
Meeker, F., Fox, D., & Weber, C. (1997). A comparison of driver behavior at railroad
grade crossings with two different protection systems. Accident Analysis &
Prevention, 29(1), 11-16.
Meeker, F. L., & Barr, R.A. (1989). An observational study of driver behavior at a
protected railroad grade crossing as trains approach*1. Accident Analysis &
Prevention, 21(3), 255-262.
Memmert, D. (2006). The effects of eye movements, age, and expertise on
inattentional blindness. Consciousness and Cognition, 15(3), 620–627.
Mesken, J., Lajunen, T., & Summala, H. (2002). Interpersonal violations, speeding
violations and their relation to accident involvement in Finland. Ergonomics,
45(7), 469-483.
Metlink Victoria. (2005). Comprehensive plan unveiled to tackle level crossing
safety. Media Release - 8th November 2005.
341
Middlestadt, S., Bhattacharyya, K., Rosenbaum, J., Fishbein, M., & Shepherd, M.
(1996). The use of theory based semistructured elicitation questionnaires:
formative research for CDC’s prevention marketing initiative. Public Health
Reports, 111(Suppl 1), 18-27.
Miller, G. A. (1965). The magical number seven plus or minus two: Some limits on
our capacity for processing information. Psychological Review, 63, 81-97.
Min, S. T., & Redelmeier, D.A. (1998). Car phone and car crashes: An ecological
analysis. Canadian Journal of Public Health, 89, 157-161.
Mitsopoulos, E., Regan, M.A., Triggs, T.J., Wigglesworth, E.C., & Tomasevic, N.
(2002). Do pictogram signs increase safety at passive crossings? A simulator
experiment. Paper presented at the 7th International Symposium on Railroad-
Highway Grade Crossing Research and Safety, Melbourne, Australia.
Mok, S., & Savage, I. (2005). Why has safety improved at rail-highway grade
crossings?. Risk Analysis, 25(4), 867-881.
Moller, H. J., Shapiro, C.M., & Kayumov, L. (2004). Effects of psychotropics on
driving performance. In M. Lader, Cardinali, D.P., & Pandi-Perumal, S.R.
(Ed.), Sleep and sleep disorders: A neuropsychopharmacological approach:
Landes Bioscience.
Mortimer, R. (1988). Human factors in highway-railroad grade crossing accidents. In
G. E. Peters & B. J. Peters (Eds.), Automotive engineering and litigation (Vol.
2, pp. 35-69). New York: Garland Law Publishing.
National Road Transport Commission. (1999). Australian road rules. Sydney: Drafted
by the Office of Legislative Drafting, Commonwealth Attorney General's
Department. Approved by the Australian Transport Council.
342
National Road Transport Commission. (2004). National heavy vehicle safety strategy:
2003-2010 (No. ISBN1 877093 33 5). Sydney: Approved by the Australian
Transport Council.
National Transportation Safety Board. (1986). Passenger/commuter trains and motor
vehicle collisions at grade crossings. Washington, DC: NTSB.
National Transportation Safety Board. (1995). Safety study factors that affect highway
fatigue in heavy truck accidents (No. PB95-917001). Springfield, VA: NTSB.
National Transportation Safety Board. (1998). Safety study: Safety at passive grade
crossings. Volume 2: Case studies. (No. PB98-917005, NTSB/SS-98/02).
Washington, D.C.: NTSB.
Nelson, T. M. (1997). Fatigue, mindset and ecology in the hazard dominant
environment. Accident Analysis & Prevention, 29, 409-415.
Netemeyer, R. G., & Burton, S. (1990). Examining the relationships between voting
behavior, intention, perceived behavioral control, and expectation. Journal of
Applied Social Psychology, 20, 661-680.
Netemeyer, R. G., Burton, S., & Johnston, M. (1991). A comparison of two models
for the prediction of volitional and goal-directed behaviors: A confirmatory
analysis approach. Social Psychology Quarterly, 54, 87-100.
Newstead, S., Cameron, M., Gantzer, S., & Vulcan, P. (1995). Modelling of some
major factors influencing road trauma in Victoria 1989–1993. Melbourne:
Monash University Accident Research Centre.
Nicolson, P., & Anderson, P. (2003). Quality of life, distress and self-esteem: A focus
group study of people with chronic bronchitis. British Journal of Health
Psychology, 8, 251-270.
343
Nicolson, P., & Anderson, P. (2003). Quality of life, distress and self-esteem: A focus
group study of people with chronic bronchitis., British Journal of Health
Psychology (Vol. 8, pp. 251-270): British Psychological Society.
NSW Government. (2006). NSW Level Crossing Strategy Council. Retrieved 4th
January, 2006, from http://www.levelcrossings.nsw.gov.au/the_lcsc.htm
NSW Roads and Traffic Authority (RTA). (2006). Community education campaigns.
Retrieved 3rd January 2006, from
http://www.levelcrossings.nsw.gov.au/campaigns.htm
Nutbeam, D., & Harris, E. (2004). Theory in a nutshell: A practical guide to health
promotion theories. London: McGraw-Hill Education.
O’Neill, B., & Mohan, D. (2002). Reducing motor vehicle crash deaths and injuries in
newly motorising countries. British Medical Journal, 324, 1142-1145.
OECD. (1990). Behavioral adaptations to changes in the road transport system.
Chicago: University of Chicago Press.
Ogden, K. W. (1996). Safer roads: A guide to road safety engineering. Sydney:
Aldershot, Avebury Technical.
Ouellette, J. A., & Wood, W. (1998). Habit and intention in everyday life: The
multiple processes by which past behavior predicts future behavior.
Psychological Bulletin, 124, 54–74.
Oxley, J., Fildes, B., Ihsen, E., Day, R., & Charlton, J. (1995). An investigation of
road-crossing behaviour of elderly pedestrians. Paper presented at the Road
Safety Research and Enforcement Conference, Fremantle, Australia.
Pack, A. I., Pack, A. M., Rodgman, E., Cucchiara, A., Dinges, D. F., & Schwab , C.
W. (1995). Characteristics of crashes attributed to the driver having fallen
asleep. Accident Analysis & Prevention, 27(6), 769-775.
344
Pajunen, K. (2002). Safety at passive grade crossings in Finland. Paper presented at
the 7th International Symposium on Rail-road Highway Grade Crossing
Research and Safety, Melbourne, Monash University.
Pajunen, K. (2004). Safety at passive grade crossings in Finland. Paper presented at
the 8th International Level Crossing Symposium, Sheffield, England.
Parker, D., & Stradling, S.G. (2001). Influencing driver attitudes and behaviour (No.
17). London: DETR.
Parker, D., . McDonald, L,. Rabbitt, P,. & Sutcliffe, P. (2000). Elderly drivers and
their accidents: The Aging Driver Questionnaire. Accident Analysis &
Prevention, 32(6), 751-759.
Parker, D., Lajunen, T., & Stradling, S. (1998). Attitudinal predictors of aggressive
driving violations. Transportation Research Part F: Traffic Psychology and
Behaviour, 1, 11– 24.
Parker, D., Manstead, A., & Stradling, S.G. (1995a). Extending the theory of planned
behavior: The role of personal norm. British Journal of Social Psychology, 34,
127–137.
Parker, D., Manstead, A.S.R., Stradling, S.G., & Reason, J.T. (1992a). Determinants
of intention to commit driving violations. Accident Analysis and Prevention,
24, 117–131.
Parker, D., Manstead, A.S.R., Stradling, S.G., Reason, J.T., & Baxter, J.S. (1992b).
Intention to commit driving violations: An application of the theory of planned
behaviour. Journal of Applied Psychology, 77(1), 94-101.
Parker, D., Reason, J. T., Manstead, A. S. R., & Stradling, S. G. (1995). Driving
errors, driving violations and accident involvement,. Ergonomics, 38(5), 1036-
1048.
345
Parker, D., Stradling, S.G., &.Manstead, A.S.R. (1996). Modifying beliefs and
attitudes to exceeding the speed limit: An intervention study based on the
theory of planned behaviour. Journal of Applied Social Psychology, 26, 1-19.
Parker, D., West, R., Stradling, S., & Manstead, A.S. (1995b). Behavioural
characteristics and involvement in different types of traffic accident. Accident
Analysis and Prevention, 27(4), 571-581.
Parker, W. (2004). Rethinking conceptual approaches to behaviour change: The
importance of context. from www.cadre.org.za
Patterson, T. L., Frith, W. J., Povey, L. J., & Keall, M. D. (2002). The effect of
increasing rural interstate speed limits in the United States. Traffic Injury
Prevention, 3, 316-320.
Perneger, T., & Smith, G.S. (1991). The driver’s role in fatal two-car crashes: A
paired “case–control” study. American Journal of Epidemiology, 134, 1138–
1145.
Pickett, W., & Grayson, G.B. (1996). Vehicle driver behaviour at level crossings (No.
98). Berkshire: Transport Research Laboratory.
Piquero, A., & Paternoster, R. (1998). An application of Stafford and Warr's
reconceptualization of deterrence to drinking and driving. Journal of Research
in Crime and Delinquency, 35(1), 3-39.
Porter, B. E., & Berry, T. D. (2001). A Nationwide Survey of Self-Reported Red
Light Running: Measuring Prevalence, Predictors, and Perceived
Consequences. Accident Analysis and Prevention, 33, 735-741.
Powles, J., & Gifford, S. (1993). Health of nations: Lessons from Victoria, Australia.
British Medical Journal,, 306, 125-127.
346
Preusser, D. F., Williams, A. F., Ferguson, S. A., Ulmer, R. G., & Weinstein, H. B.
(1998). Fatal crash risk for older drivers at intersections. Accident Analysis &
Prevention, 30, 151-159.
Public Transport Safety Victoria, D. o. I. (2005). Study of road user behaviour at
Springvale Road, Nunawading railway level crossing. Melbourne: Department
of Infrastructure.
Queensland Transport. (2007). Fatigue management. Retrieved 15th May, from
http://www.transport.qld.gov.au/Home/Safety/Road/Heavy_vehicles/Fatigue_
management/
Quinlan, M. (2001). Report of inquiry into safety in the long haul trucking industry.
Sydney: University of New South Wales.
Rafaely, V., Meyer, J., Zilberman-Sandler, I., & Viener, S. (2006). Perception of
traffic risks for older and younger adults. Accident Analysis & Prevention,
38(6), 1231–1236.
Ragland, D. R., Satariano, W. A., & MacLeod, K. E. (2004). Reasons given by older
people for limitation of avoidance of driving. Gerontologist, 44(2), 237–244.
Rail Safety and Standards Board. (2004). Road vehicle level crossings special topic
report. London: Rail Safety and Standard Board.
Rajalin, S. (1994). The connection between risky driving and involvement in fatal
accidents. Accident Analysis & Prevention, 26, 555–562.
Ramsay, E., & Prem, H. (2000). Development of an Austroads Heavy Vehicle
Nomenclature System: Discussion Paper. Sydney: Austroads.
Ranney, T. A., Mazzae, E., Garrott, R., & Goodman, M.J. (2000). Driver distraction:
Past, present and future. from http://www-nrd.nhtsa.dot.gov/departments/nrd-
13/driver-distraction/PDF/233.PDF
347
Rapoza, A. S., & Fleming, G.G. (2002). Determination of a sound level for railroad
horn regulatory compliance. Washington: U.S. Department of Transportation,
Federal Railroad Administration.
Reason, D., Manstead, A., Stradling, S., Baxter, B., & Campbell, K. (1990a). Errors
and violations on the road: A real distinction? Ergonomics, 33, 1315–1332.
Reason, J. (1997). Managing the risks of organizational accidents. Brookfield:
Ashgate.
Reason, J., Manstead, A., Stradling, S., Baxter, J., & Campbell, K. (1990b). Errors
and violations on the roads: A real distinction? Ergonomics., 33(10-11), 1315-
1332.
Reason, J. T. (2000). Grace under fire: compensating for adverse events in
cardiothoracic surgery. Paper presented at the 5th Conference on Naturalistic
Decision Making, Tammsvik, Sweden.
Rechnitzer, G. (2002). Innovative measures to reduce crash occurrence and severe
injury risk at passive railroad crossings. Paper presented at the Seventh
International Symposium on Railroad-Highway Grade Crossing Reserach and
Safety, Melbourne, Australia.
Regan, M. Y., K. (2003). Driver distraction: A review of the literature and
recommendations for countermeasures. Paper presented at the Road Safety
Research, Policing and Education Conference, Melbourne.
Rhodes, R. E., & Courneya, K. S. (2004). Threshold assessment of attitude, subjective
norm, and perceived behavioural control for predicting exercise intention and
behaviour. Psychology of Sport and Exercise, Article in press.
Richards, S. H. (1990). Human factors issues affecting driver behavior at rail–
highway grade crossings with active traffic control. Paper presented at the
348
International Symposium on Railroad–Highway Grade Crossing Research and
Safety, University of Tennessee.
Richter, E. D., Barach, P., Friedman, L., Krikler, S., & Israeli, A. (2004). Raised
speed limits, speed spillover, case-fatality rates and road deaths in Israel: A 5-
year follow-up. American Journal of Public Health, 94(4), 568-574.
Risser, R., & Nickel, W.-R. (2004). Theories of science in traffic psychology. In R. T.
& R. Huguenin (Eds.), Traffic and Transport Psychology. Amsterdam:
Elsevier.
Ronis, D. L., Yates, J.F., & Kirscht, J.P. (1989). Attitudes, decisions, and habits as
determinants of repeated behavior. In A. R. Pratkanis, S.J. Breckler, & A.G.
Greenwald, (Ed.), Attitude structure and function (pp. 213–239). Hillsdale,
NJ: Lawrence Erlbaum Associates, Inc.
Rosenbloom, T., & Wolf, Y. (2002). Sensation seeking and detection of risky road
signals: A developmental perspective. Accident Analysis and Prevention, 34,
569–580.
Rosenstock, I. (1966). Why people use health services. Milbank Memorial Fund
Quarterly, 44(94-124).
Rosenstock, I. M. (1974). Historical origins of the Health Belief Model. Health
Education Monographs, 2, 1-8.
Rosenstock, I. M., Strecher, V.J., & Becker, M.H. (1994). The Health Belief Model
and HIV Risk Behavior Change. In J. Peterson & R. DiClemente (Eds.),
Preventing AIDS: Theory and Practice of Behavioral Interventions (pp. 5-24.).
New York: Plenum Press.
Rosenthal, R. (1991). Meta-analytic procedures for social research: Applied social
research methods series (Volume 6). London: SAGE Publications.
349
Rothengatter, T. (1991). Automatic policing and information systems for increasing
traffic law compliance. Journal of Applied Behaviour, 24(1), 85–87.
Rothman, A., Klein, W., & Weinstein, N. (1996). Absolute and relative biases in
estimations of personal risk. Journal of Applied Social Psychology, 26, 1213-
1236.
Royal, D. (2002). National survey of distracted and drowsy driving attitudes and
behaviours. Washington, DC: U.S. Department of Transportation, NHTSA.
Russell, E. R. (1974a). Analysis of driver reaction to warning devices at a high-
accident rural grade crossing: Purdue University/Indiana State Highway
Commission.
Russell, E. R. (1974b). Analysis of driver reaction to warning devices at a high
accident rural grade crossing. (No. #74-16): Purdue University/Indiana State
Highway Commission.
Russell, E. R., & Kent, W. D. (1993). Two human factors studies to evaluate new
passive devices at rail– highway grade crossings and a comparison to
traditional traffic study methods. Paper presented at the International
Symposium on Railroad–Highway Grade Crossing Research and Safety,
University of Tennessee.
Russell, E. R., & Konz, S. (1980). Night visibility of trains at railroad-highway grade
crossings. Transportation Research Record, 773, 7-11.
Russo, F. A., Lantz, M.E., English, G.W., & Cuddy, L.L. (2003, July 6-9). Increasing
effectiveness of train horns without increasing intensity. Paper presented at the
2003 International Conference on Auditory Display, Boston.
350
Rutter, D., Quine, L., & Arnold, L. (1998). Predicting and understanding safety
helmet use among schoolboy cyclists: A comparison of the Theory of Planned
Behaviour and the Health Belief Model. Psychology and Health, 13, 251-269.
Ryan, G. A., Legge, M., & Rosman, D. (1998). Age related changes in drivers’ crash
risk and crash type. Accident Analysis & Prevention, 30(3), 379–387.
Salazar, M. K. (1991). Comparison of four behavioural theories. American
Association of Occupational Health Nurses Journal, 39(3), 128-135.
Sampson, P. (1972). Qualitative research and motivation research. In R. M. Worcester
(Ed.), Consumer Market Research Handbook (pp. 7-27). London: McGraw-
Hill.
Sanders, J. H. (1976). Driver performance in countermeasure development at railroad-
highway grade crossings. Transportation Research Record, 28-37.
Sanne, J. M. (1999). Creating safety in air traffic control. Lund, Sweden: Arkiv.
Savage, I. (2006). Does public education improve rail-highway crossing safety?
Accident Analysis & Prevention, 38(2), 310-316.
Schifter, D., & Ajzen, I. (1985). Intention, perceived control, and weight loss: An
application of the theory of planned behaviour. Journal of Personality and
Social Psychology, 49, 843-851.
Schmid, F., & Watson, C. (2004). A worldwide comparison of level crossing
accidents and a formula for predicting accident rates. Paper presented at the
8th International Level Crossings Symposium and Managing Trespass
Seminar, Sheffield, England.
Schoppett, D. W., & Hoyt, D. W. (1968). Factor influencing safety at highway-rail
grade crossings. (No. 50, NCHRP). McLean, VA: A.M. Voorhees and
Associates.
351
Sentinella, J. (2004). Guidelines for evaluating road safety education interventions.
London: Department for Transport.
Severin, W. J., & Tankard, J.W.Jnr. (2001). Theories of persuasion. In
Communication theories (5th ed.). New York: Longman.
Shaheen, S. A., & Niemeier, D. A. (2001). Integrating vehicle design and human
factors: Minimizing elderly driving constraints. Transportation Research Part
C, 9, 155-174.
Sheeran, P., Milne, S., Webb, T.L., & Gollwitzer, P.M. (2004). Implementation
intentions and health behaviours. In M. Conner, & Norman, P. (Ed.),
Predicting health behaviour: Research and practice with social cognition
models. Buckingham: Open University Press.
Sheeran, P., Trafimow, D., & Armitage, C.J. (2003). Predicting behaviour from
perceived behavioural control: Tests of the accuracy assumption of the theory
of planned behaviour. British Journal of Social Psychology, 42, 393-410.
Sheppard, B. H., Hartwick, J., & Warshaw, P. R. (1988). The theory of reasoned
action: A meta-analysis of past research with recommendations for
modifications and future research. Journal of Consumer Research, 15, 325-
343.
Sheppard, B. H., Hartwick, J., & Warshaw, P. R. (1988). The theory of reasoned
action: A meta-analysis of past research with recommendations for
modifications and future research. Journal of Consumer Research, 15, 325-
343.
Shier, R. (2004). Statistics: Paired t-tests. Retrieved 3rd May 2007, from
http://www.site.uottawa.ca/~welazmeh/teaching/CSI7163/Pairedttest.pdf
352
Shinar, D., & Raz, S. (1982). Driver response to different railroad crossing protection
systems. Ergonomics, 25(9), 801-808.
Shinar, D., & Scheiber, F. (1991). Visual requirements for safety and mobility of
older drivers. Human Factors, 33(5), 507-519.
Silock, D., Smith, K., Knox, D., & Beuret, K. (1999). What limits speed? Factors that
affect how fast we drive. Basingstoke: AA Foundation for Road Safety
Research.
Simms, B. (1985). Perception and driving: Theory and practice. Occupational
Therapy, 363-366.
Simons, D. (2000). Attentional capture and inattentional blindness. Trends in
Cognitive Science, 4(4), 147-155.
Simpson, H. M. (1995). New to the road: Reducing the risks for young motorists.
Paper presented at the First Annual International Symposium of the Youth
Enhancement Service, Los Angeles, CA.
Simsekoglu, Ö., & Lajunen, T. (2004). Theory of planned behaviour health belief
model and seat belt use among Turkish passengers. Paper presented at the 3rd
International Conference on Traffic and Transport Pyschology, Nottingham,
England.
Single, E., & Rohl, T. (1997). The national drug strategy: Mapping the future: An
evaluation of the national drug strategy 1993-1997. Canberra: AGPS.
Sjöberg, L., Moen, B.-E., & Rundmo, T. (2004). Explaining risk perception: An
evaluation of the psychometric paradigm in risk perception research.
Trondheim, Norway: Norwegian University of Science and Technology.
Sjogren, H., Bjornstig, U., Eriksson, A., Ohman, U., & Solarz, A. (1997). Drug and
alcohol use among injured motor vehicle drivers in Sweden: prevalence,
353
driver, crash, and injury characteristics. Alcoholism: Clinical and
Experimental Research,, 21, 968–973.
Sliogeris, J. (1992). 110 kilometre per hour speed limit – Evaluation of road safety
effects. (No. GR92-8). Melbourne, Australia: VicRoads.
South, D. (1998). General deterrence and behaviour change: A comment on the
Australian Psychological Society position paper on punishment and behaviour
change. Australian Psychologist, 33(1), 76-78.
Stafford, M. C., & Warr, M.,. (1993). A reconceptualization of general and specific
deterrence. Journal of Research in Crime and Delinquency, 30(2), 123-135.
Standards Australia. (1993). Manual of uniform traffic control. Part 7: Railway
crossings. Homebush: Standards Australia.
Standards Australia. (2007). Manual of uniform traffic control devices. Part 7:
Railway crossings (AS 1742.7—2007). Sydney: Standards Australia.
Stephen, N. (2002). Some legal considerations. Paper presented at the 7th
International Level Crossing Symposium, Melbourne, Australia.
Stott, P. F. (1987). Automatic open level crossings: A review of safety. London:
HMSO.
Stutts, J., Stewart, J. R., & Martell, C. (1998). Cognitive test performance and crash
risk in an older driver population. Accident Analysis and Prevention, 30(3),
337– 346.
Sudman, S. (1976). Applied sampling. New York: Academic Press.
Sullman, M. J., Meadows, M.L., & Pajo, K.B. (2002). Aberrant driving behaviours
amongst New Zealand truck drivers. Transportation Research Part F, 5, 217–
232.
354
Summala, H., & Mikkola, T. (1994). Fatal accidents among car and truck drivers:
Effects of fatigue, age, and alcohol consumption. Human Factors, 36, 315-
326.
Sutton, S. (1993). The past predicts the future: Interpreting behaviour-behaviour
relationships in social-psychological models of health behaviours. In D. R.
Rutter & L. Quin (Eds.), Social psychology and health: European perspectives
(pp. 78-88). Avesbury.: Aldershot.
Sutton, S., Eiser, R., & Richard, J. (1990). The decision to wear a seat belt: The role
of cognitive factors, fear and prior behaviour. Psychology and Health, 4, 111-
123.
Swann, P. (2002). The major issues of drugs, alcohol and fatigue in heavy vehicle
safety. Paper presented at the National Heavy Vehicle Safety Seminar,
Melbourne.
Sweatman, P., Ogden, K. W., Haworth, N., Corben, B., Rechnitzer, G., &
Diamantopoulou, K. (1995). Heavy vehicle crashes in urban areas (No. CR
155). Canberra: Federal Office of Road Safety.
Swedish Road and Traffic Research Institute. (1981). Research Program 1981-1985.
Linkoping, Sweden: Swedish Road and Traffic Research Institute (VTI).
Sweedler, B. M. (1997). The worldwide decline in drinking and driving – where are
we now. Paper presented at the 14th International Conference on Alcohol,
Drugs and Traffic Safety, Annecy, France.
SWOV Institute for Road Safety Research. (2005). Cost-benefit analysis of road
safety measures. Leidschendam, the Netherlands.
355
Tay, R. (2004). The relationship between public education and enforcement
campaigns and their effectiveness in reducing speed related serious crashes.
International Journal of Transport Economics, 31(2), 251-255.
Tay, R. (2005). The effectiveness of enforcement and publicity campaigns on serious
crashes involving young male drivers: Are drink driving and speeding similar?
Accident Analysis and Prevention, 37, 922–929.
Taylor, S., & Todd, P. (1995). Decomposition and crossover effects in the theory of
planned behavior: A study of consumer adoption intentions. International
Journal of Research in Marketing, 12, 137-156.
The Health Communication Unit. (2002). Using focus groups. Toronto, Ontario:
Centre for Health Promotion, University of Toronto.
Thiffault, P. B., J. (2003). Fatigue and individual differences in monotonous
simulated driving. Personality and Individual Differences,, 34(1), 159-179.
Tilley, A. (1990). An introduction to psychological research and statistics (2nd ed.).
Brisbane: Pineapple Press.
Tillman, W. A., & Hobbs, G.E. (1949). The accident-prone automobile driver. A
study of the psychiatric and social background. American Journal of
Psychiatry, 106, 321-331.
Toepell, A. (2003). The health belief model and safer sex: Implications for women’s
health. Women's Health and Urban Life: An International and
Interdisciplinary Journal, 2(1), 22-41.
Trafimow, D., & Fishbein, M. (1994). The importance of risk in determining the
extent to which attitudes affect intentions to wear seat belts. Journal of
Applied Social Psychology, 24(1), 1-11.
356
Transportation Development Centre. (2003). Locomotive horn evaluation:
Effectiveness at operating speeds. Ottawa: Transport Canada.
Transportation Research Board. (1998). Managing speed: Review of current practice
for setting and enforcing speed limits. (No. 254). Washington DC.:
Transportation Research Board.
Trbovich, P., & Harbluk, J. L. (2003). Cell phone communication and driver visual
behavior: The impact of cognitive distraction. Paper presented at the
Conference on Human Factors in Computing Systems, Fort Lauderdale,
Florida.
Treat, J. R., Tumbas, N.S., McDonald, S.T., Shinar, D., Hume, R.D., Mayer, R.D.,
Stansifer, R.L. & Castallen, N.J. (1977). Tri-level study of the causes of traffic
accidents: Final report. Washington, D.C.: NHTSA.
Triandis, H. C. (1977). Interpersonal behavior. Monterey, CA: Brooks/Cole.
Triandis, H. C. (1980). Values, attitudes, and interpersonal behavior. Nebraska
Symposium on Motivation, 27, 195-259.
Triggs, T. J., & Smith, K.B. (1998). Young driver research program: Digest of
reports and principal findings of the research. Canberra: Federal Office of
Road Safety.
Turner, C., & McClure, R. (2004). Quantifying the role of risk-taking behaviour in
causation of serious road crash-related injury. Accident Analysis and
Prevention, 36, 383–389.
Turnock, B. J. (1997). Public Health: What it is and how it works? Gaithersburg ,
USA: Aspen Publishers.
357
Tustin, B. H., Richards, H., McGee, H. & Patterson, R. (1986). Railroad-highway
grade crossing handbook (2nd ed.). Virginia: U.S. Department of
Transportation.
United Nations. (2000). Evaluation of cost-effective systems for railway level-crossing
protection. New York: Economic and Social Commission for Asia and the
Pacific.
van Belle, G., Meeter, D., & Farr, W. (1975). Influencing factors for railroad-highway
grade crossing accidents in Florida. Accident Analysis and Prevention, 7, 103-
112.
Van den Putte, B. (1991). 20 years of the theory of reasoned action of Fishbein and
Ajzen: A meta-analysis. University of Amsterdam, Amsterdam.
van der Horst, R., & Bakker, P. (2002). The effectiveness of safety measures at
railway level crossings on road user behaviour. Paper presented at the 15th
ICTCT workshop, Brno, Czech Republic.
Vastine, A., Gittelsohn, J., Ethelbah, B., Anliker, J., & Cabellero, B. (2005).
Formative research and stakeholder participation in intervention development.
American Journal of Health Behaviour, 29(1), 57-69.
VicRoads. (2000). Road safety strategy for Victoria 2000-2005. Melbourne,
Australia: VicRoads.
Victoria Police State Highway Task Force. (1995). Four case studies and results of
Operation Tall Poppies: Inquiry into the effects of drugs (other than alcohol)
on road safety in Victoria. Melbourne.: Parliamentary Road Safety
Committee.
358
Vingilis, E. (1990). A new look at deterrence. In R. J. Wilson & R. E. Mann (Eds.),
Drinking and driving: Advances in research and prevention. New York:
Guilford Press.
Wallace, A., Ibrahim, N., Davey, J., Hughes, B., & Ford, G. (2006). Motorist
behaviour at railway level crossings: An exploratory study of experiences of
train drivers. Paper presented at the Conference On Railway Engineering
(CORE), Sydney.
Wallace, B. (2003). External-to-vehicle driver distraction. Edinburgh: Scottish
Executive Social Research.
Walton, D. (1999). Examining the self-enhancement bias: Professional truck drivers’
perceptions of speed, safety, skill and consideration. Transport Research Part
F, 2(2), 91–113.
Wang, B., Hensher, D., & Ton, T. (2002). Safety in the road environment: A driver
behavioural response perspective. Transportation, 29(3), 252-270.
Ward, N. J., & Wilde, G. J. S. (1995). A comparison of vehicular approach speed and
braking between day and nighttime periods at an automated railway crossing.
Safety Science, 19, 31-44.
Ward, N. J., & Wilde, G. J. S. (1996). Driver approach behaviour at an unprotected
railway crossing before and after enhancement of lateral sight distances: An
experimental investigation of a risk perception and behavioural compensation
hypothesis. Safety Science, 22(1-3), 63-75.
Washington State Department of Transportation. (2004). What are rumble strips?
Retrieved 14 December, 2004, from
http://www.wsdot.wa.gov/EESC/Design/policy/RumbleStripWeb/WhatAreRu
mbleStrips.htm
359
Watson, B. (2004a). How effective is deterrence theory in explaining driver
behaviour: A case study of unlicensed driving. Paper presented at the Road
Safety Research, Policing and Education Conference, Perth, W.A.
Watson, B. (2004b). The psychological characteristics of on road behaviour of
unlicensed drivers. Queensland University of Technology, Brisbane.
Weale, R. A. (1992). The senescence of human vision. Oxford: Oxford University
Press.
Webb, T. L., & Sheeran, P. (2004). Integrating concepts from goal theories to
understand the achievement of personal goals. (Manuscript under review.).
Weinreich, N. (1999). Hands-on social marketing: A step-by-step guide. London:
Sage Publications.
Wells-Parker, E., Kenne, D., Spratke, K., & Williams, M. (2000). Self-efficacy and
motivation for controlling drinking and drinking/driving: an investigation of
changes across a driving under the influence (DUI) intervention program and
of recidivism prediction. Addictive Behaviours, 25(2), 229-238.
Wengraf, T. (2001). Qualitative research interviewing. London: SAGE.
West, C., Gildengorin, G., Haegerstrom-Portnoy, G., Lott, L., Schneck, M., &
Brabyn, J. (2003). Vision and driving self-restriction in older adults. Journal
of the American Geriatrics Society, 51, 1348-1355.
West, R. J. (1995, 4–5 September). Individual differences in accident risk: A review of
findings and an examination of methods. Paper presented at the Behavioural
research in road safety VI., London.
White, M., Walker, J., Glonek, G., & Burns, N. (2000). Reinvestigation of the
effectiveness of the Victorian TACs road safety campaign. Adelaide: Transport
South Australia.
360
Wigglesworth, E. (1979). The epidemiology of road-rail crossing accidents in
Victoria, Australia. Journal of Safety Research, 11(4), 162-171.
Wigglesworth, E. (2001). A human factors commentary on innovations at railroad-
highway grade crossings in Australia. Journal of Safety Research, 32, 309-
321.
Wigglesworth, E. (2002). Passive railroad-highway grade crossings: What are the
research priorities? Paper presented at the 7th International Symposium on
Railroad-Highway Grade Crossing Research and Safety, Monash University,
Melbourne.
Wigglesworth, E., & Uber, C. (1991). An evaluation of the railway level crossing
boom barrier program in Victoria, Australia. Journal of Safety Research,
22(3), 133-140.
Wigglesworth, E. C. (1978). The fault doctrine and injury control. Journal of Trauma,
18, 789.
Wilde, G. J. S. (1982). The theory of risk homeostasis: Implications for safety and
health. Risk Analysis, 2, 209–258.
Wilde, G. J. S. (2001). Target risk 2: A new psychology of safety and health. Toronto:
PDE Publications.
Wilde, G. J. S., Hay, M. C., & Brites, J. N. (1987). Video-recorded driver behaviour
at railway crossings: Approach speeds and critical incidents. (No. 87-61).
Kingston: Canadian Institute for Guided Ground Transport at Queen’s
University.
Williams, D., & Creber, G. (2005). Level crossings in NSW - Assessing safety and
prioritising works. Paper presented at the IPWEA NSW Division Annual
Conference, Sydney.
361
Williamson, A. (2003a). Why are young drivers over represented in crashes?
Summary of the issues. Update of literature review: Literature 2000 to 2003.
Sydney, Australia: Motor Accidents Authority.
Williamson, A., Irvine, P., & Friswell, R. (2003b). What is the involvement of heavy
trucks in crashes in NSW? Paper presented at the Road Safety Research,
Policing and Education Conference, Sydney.
Winter, D. (1988). Older drivers – their perception of risk. Effects of aging on driver
performance. Paper presented at the Passenger Car Meeting and Exposition,
Michigan.
Witte, K., & Donohue, W.A. (2000). Preventing vehicle crashes with trains at grade
crossings: The risk seeker challenge. Accident Analysis & Prevention, 32(1),
127-139.
Wittenbraker, J., Gibbs, B.L., & Kahle, L.R. (1983). Seat belt attitudes, habits and
behaviors: An adaptive amendment to the Fishbein Model. Journal of Applied
Social Psychology, 13(5), 406-421.
Wong, S. C., Leung, B. S. Y., Loob, B. P. Y., Hung, W. T., & Lo, H. K. (2004). A
qualitative assessment methodology for road safety policy strategies. Accident
Analysis & Prevention, 36(2), 281-293.
Wood, J. (2002). Aging, driving and vision. Clinical and Experimental Optometry,
85(4), 214-220.
Woods, D. D., Johannesen, L.J., Cook, R.I., & Sarter, N.B. (1994). Behind human
error: Cognitive systems, computers and hindsight. Dayton: CSERIAC.
Woolley, J. (2000). In-car driver Training at High schools: A Literature Review (No.
6/2000). Adelaide: Transport SA.
362
World Health Organization. (2004). World report on road traffic injury prevention.
Geneva: World Health Organization.
Young, C. (1990). How serious is the ageing problem and what can be done about it?
Working Papers in Demography (No. 18). Canberra: Australian National
University.
Zador, P. (2003). Analysis of the safety impact of train horn bans at highway-rail
grade crossings. Washington DC: Federal Railroad Administration (FRA).
Zalinger, D. A., Rodgers, B. A., & Johri, H. P. (1977). Calculation of hazard indices
for highway-railway crossings in Canada. Accident Analysis and Prevention,
9, 257-273.
Zimring, F., & Hawkins, G. (1976). Deterrence: A review of the literature. In E.
Fattah (Ed.): Law Reform Commission of Canada.
Zuckerman, M. (1994a). Behavioral expressions and biosocial bases of sensation
seeking. New York: Cambridge Press.
Zuckerman, M. (1994b). Behavioural expressions and biological bases for sensation
seeking. Cambridge: University of Cambridge Press.
363
APPENDICES
364
Appendix 1: Review of models/theories
365
THEORY OF PLANNED BEHAVIOUR
Extension to the theory of reasoned action
The theory of planned behaviour (Ajzen, 1985), an extension of the theory of
reasoned action (Fishbein, 1967), is one model that has been successful in predicting
behaviour in road safety. This theory has previously been used to understand
pedestrians’ road crossing decisions and predicting their intentions (Evans, 2003,
Evans, 1998), to examine intention to commit driving violations (Parker, 1992b), as
well as modifying beliefs and attitudes to exceeding the speed limit (Parker, 1996)
and driver’s compliance with speed limits (Elliott, 2003b). Although this theory has
been contentious in its ability to predict behaviour, there is no doubt that this theory
embodies an important perspective from which to examine level crossing behaviour.
The theory of reasoned action (Fishbein, 1967, Fishbein, 1975) is one of the
most widely studied social psychology models of attitude and behaviour. This theory
developed in 1967, includes attitude as one of the important factors influencing
behavioural intention (Severin, 2001) and assumes that people are rational in making
decisions based on the information available to them. According to Ajzen and
Fishbein (1980), a person will normally act in accordance with their intentions
although a perfect correlation between intention and behaviour does not exist. The
figure below illustrates the theory of reasoned action (Ajzen and Fishbein, 1980).
Theory of Reasoned Action
Subjective
Norm
Attitude
Toward the Behaviour
Behavioural
Intention
Behaviour
366
The theory of reasoned action suggests that a person’s behaviour is determined
by their intention to perform the behaviour and that this intention is a function of their
attitude towards the behaviour as well as their subjective norm (Ajzen and Fishbein,
1980). Attitude is defined as being the “affective and instrumental evaluations of
performing the behavior by the individual”(Rhodes and Courneya, 2004, p2) while
subjective norm is defined as the “social pressure on the individual to perform or not
to perform a particular behaviour” (Rhodes and Courneya, 2004, p2). The
determinants of the attitudinal and normative constructs (i.e. information people have
about themselves and the world in which they live) are important in influencing
behaviour (Ajzen and Fishbein, 1980). Beliefs are underlying to a person’s attitudes
and subjective norms, and ultimately determine behavioural intention as well as
behaviour (Ajzen and Fishbein, 1980). According to Fishbein (1967) this theory
seeks not only to predict human behaviour but also understand it, thus making it
important to identify and analyze the determinants of intentions.
Fishbein and Ajzen’s (1980) theory of reasoned action has received strong
empirical support over the years (Sheppard, 1988), however many attempts have been
made to improve its predictability, either by adding new variables or by making
changes to its internal structure (Bagozzi, 1992). In 1985, Ajzen extended the TRA
and proposed an additional variable, ‘perceived behavioural control’ to the model to
account for the limitation in dealing with behaviours over which people have
incomplete volitional control (Taylor, 1995). According to Taylor and Todd (1995),
perceived behavioural control is the person’s control beliefs weighted by the
perceived facilitation of the control factor that either inhibits or facilitates the
behaviour. Perceived behavioural control is also regarded as a co-determinant of
behaviour, although it is largely dependent on the accuracy of people’s perceptions of
control (Ajzen, 1991, Sheeran, 2003). Studies have indicated that measurements of
perceived behavioural control improve predictions of intention from attitude and
subjective norm as well as predictions of behaviour from intention (Ajzen, 1991,
Netemeyer et al., 1991, Schifter and Ajzen, 1985, Gärling, 1992). This additional
variable accounts for situations where a person has less than complete control over
their behaviour and may be relevant for level crossing behaviour (Taylor, 1995).
367
Principles and overview
The theory of planned behaviour (Ajzen, 1985) was developed out of the need
to deal with the limitations of the theory of reasoned action in terms of behaviours in
which people have incomplete volitional control (Doll, 1992). According to the
theory of planned behaviour (see Figure 17), behaviour can be predicted from a
person’s intentions to perform the behaviour and from their perceptions of control
over the behaviour. This theory proposes that behaviour intentions are the main
determinant of behaviour (Elliott, 2003b) and intention is assumed to be influenced by
perceived behavioural control in addition to attitude and subjective norm (Ajzen and
Madden, 1986, Gärling, 1992, Netemeyer, 1990). By considering these influences,
this theory is “in this way extended to include the prediction of a broader class of
behaviours than those which are under volitional control” (Gillholm, 1996, p3).
Fundamentally, this theory hypothesizes three conceptually independent determinants
of intention:
• Attitude toward the behaviour which refers to the degree to which a person has
a favorable or unfavorable evaluation or appraisal of the behaviour in
question;
• Subjective norm (social factor) which refers to the perceived social pressure
to perform or not to perform the behaviour; and
• Perceived behavioural control which refers to the perceived ease or difficulty
of performing the behaviour and is assumed to reflect past experience as well
as anticipated impediments and obstacles.
(Doll, 1992)
Attitudes are a person’s overall evaluation of the behaviour (e.g. “it is good to
stop at a level crossing when the flashing lights are activated”). Recent evidence
suggests that it is important to differentiate affective from instrumental attitudes in
relation to behaviours, which might realistically be considered as having an affective
element (Lawton, 1997, Parker, 1998, Parker, 1996). Instrumental attitudes are
related to benefits (e.g. “it is beneficial to obey warning systems at level crossings”)
and disadvantages (e.g. “it takes up time waiting for a train to pass”), while emotional
attitudes refer to emotions related to safety at level crossings (e.g. “not proceeding
368
through a level crossing when the train has passed but the flashing lights are still
activated, makes me feel silly”).
The role of beliefs in the theory of planned behaviour is that at the most basic
level of explanation, behaviour is a function of salient information or beliefs, relevant
to the behaviour (Doll, 1992). Subjective norms consist of a person’s beliefs about
whether significant others (e.g. spouse, colleagues, friends) think that he or she should
engage the behaviour (Conner, 1996). According to Miller (1965), people can hold a
great many beliefs about any given behaviour, but they can concentrate on only a
relatively small number at any given moment. As such, it is these salient beliefs that
are thought to be the main determinant of a person’s intentions and actions (Doll,
1992). Doll and Ajzen (1992) propose that there are three types of salient beliefs:
behavioural beliefs (assumed to influence attitudes towards the behaviour); normative
beliefs (constitute the underlying determinants of subjective norms); and control
beliefs (provide the basis for perceptions of behavioural control).
Perceived behavioural control forms the third variable or predictor of intentions.
This refers to a person’s perception of the extent to which performance of the
behaviour is easy or difficult (Conner, 1996). As such, in the theory of planned
behaviour, perceived behavioural control has both direct and mediated (by
behavioural intention) effects on behaviour (Lajunen, 2004). According to Ajzen
(1991) and Sheeran, Trafimow and Armitage (2003), perceived control is (with
intention) also considered a co-determinant of behaviour, although perceived control-
behaviour relationship is dependent on the precision of people’s perceptions of
control. The concept of perceived behavioural control is not original to the theory of
planned behaviour as it also appears in the health belief model (Rosenstock, 1966)
where it is termed barriers, and also in the model of interpersonal behaviour (Triandis,
1977) where it takes the form of facilitating conditions (Ajzen, 2002a). However,
Bandura’s work on self-efficacy is perhaps the originator of perceived behavioural
control (Bandura, 1977a, Bandura, 1989, Bandura, 1997). Like attitude and subjective
norm, perceived behavioural control can be measured by asking direct questions to a
person about their capability to perform a behaviour or indirectly on the basis of
beliefs about ability to deal with specific inhibiting or facilitating factors (Ajzen,
2002a).
369
Theory of Planned Behaviour
Application to level crossing behaviour
The theory of planned behaviour has been applied to driver behaviour and
efforts to predict driver violations on this basis, are delivering initial results (Parker,
1995, Parker, 1992b, Parker, 1992a). Given that there is strong support from studies
on drivers’ compliance with speed limits (Elliott, 2003b), it seems realistic that
desirable changes in drivers’ attitudes, subjective norms, and perceptions of control
may lead to consequent changes in their intentions and behaviour. However, previous
studies examining driver behaviour have emphasized the importance of distinguishing
between aberrant behaviour constituting (deliberate) violations and behaviour
resulting from errors and lapses (Reason, 1990).
With regards to driver behaviour at level crossings, Pickett and Grayson (1996)
suggest that there are three main reasons for non-observance of signals at active
crossings:
• Drivers who are unwilling to stop because they believe they have plenty of
time to cross before the train arrives;
• Drivers who are unable to stop because they are too close to the stop line at the
Subjective
Norm
Attitude Toward the Behaviour
Behavioural
Intention
Behaviour
Perceived Behavioural
Control Actual Behavioural
Control
Behavioural
Beliefs
Normative
Beliefs
Control Beliefs
370
onset of flashing lights, or because someone is driving too close behind; and
• Drivers who are unaware of the signals because they are inattentive or are
distracted.
According to Pickett and Grayson (1996) drivers may base their decision to
cross at a level crossing on their previous experience of either the same crossing
(familiarity) or of other crossings (association). When drivers are exposed to the
same phenomenon repeatedly, as in the concept of ‘mental set’, they come to expect it
(Pickett, 1996). Additionally, drivers familiar with one level crossing may transfer
their experience to a new crossing, and be complacent in the new situation (Pickett,
1996). Both of these situations may lead to drivers’ non-observance of warning and
protection systems. However, findings from Pickett and Grayson’s (1996) British
study involving analysis of a sample of witness statements obtained by the British
Transport Police during a campaign on red signal violations at level crossings,
indicates that the majority of signal violations at level crossings are intentional and
deliberate.
Evidence from fatal accidents in Victoria, Australia contradicts this British
study on red signal violations. In 1979, Wigglesworth examined 85 consecutive fatal
crashes involving motor vehicles and trains at both active and passive level crossings,
with data being taken from detailed police reports prepared for the coroner. This
study indicated that at least 86% of drivers killed in the level crossing crash lived
locally and were therefore familiar with the existence of this crossing (Wigglesworth,
1979). This Victorian evidence suggests that the drivers killed at level crossings were
not violating warning or protection systems, but rather law-abiding citizens going
about their daily work and attributable to human overload unrelated to any breach of
regulation (Wigglesworth, 2001). Therefore, it would seem that from these Victorian
coroner’s reports, that the category that fatalities generally lie in is “those who are
unaware of the signals because they are inattentive or are distracted”. In accordance
with previous studies emphasizing the importance of distinguishing between aberrant
behaviour constituting (deliberate) violations and behaviour resulting from errors and
lapses (Reason, 1990), it appears that these fatal accidents are a result of the latter.
While evidence suggests that drivers that violate road rules are more likely to be
crash-involved (Parker), it is unknown if this can be translated to level crossing
crashes.
371
Criticisms of the theory of planned behaviour
Although many health behaviour studies have found that intentions predict
corresponding behaviour quite well, there has long been a concern by researchers that
there are observed discrepancies between these two constructs (Campbell, 1963,
Blumer, 1955). Sheeran et al. (2004) propose that there are three processes that
underlie this discrepancy between intention and behaviour. Intention viability is the
first process that is defined as being “impossible for most decisions to find expression
in the absence of particular abilities, resources, or opportunities” (Sheeran, 2004, p3).
The second process is intention activation refers to the “extent to which contextual
demands alter the salience, direction, or intensity of a focal intention relative to other
intentions” (Sheeran, 2004, p4). The third process suggested by Sheeran et al. (2004)
concerns intention elaboration in which people may “fail to engage in, or elaborate in
sufficient detail, an analysis of the particular actions and contextual opportunities that
would permit realization of their intention”(Sheeran, 2004, p5).
The original derivation of the theory of planned behaviour (Ajzen, 1985) does
not define intention in relation to actual performance but rather defines intention in
terms of trying to perform a given behaviour. Early work on this theory however,
showed strong correlations between measures of the model’s variables that raised
questions about trying to perform a given behaviour and measures that dealt with the
actual performance of the behaviour (Ajzen and Madden, 1986). However, according
to Doll and Ajzen (1992) the relative importance of intention and perceived
behavioural control in the prediction of behaviour is expected to vary across situations
and across different behaviours.
HEALTH BELIEF MODEL
Origins and overview
The health belief model (HBM) developed by Rosenstock (1974), is one of the
most widely used conceptual frameworks for understanding health behavior in terms
of individual decision-making. This model which was developed during the 1950s
and influenced by works of social psychologist Kurt Lewin, was originally designed
372
to systematically explain and predict preventative health behaviour and focused on the
relationship between health behaviours, practices and utilization of health services
(Rosenstock, 1974). This model has been applied to numerous public health issues
including breast-self examination, immunization, sexual risk behaviours and the
transmission of infectious diseases.
One of the major assumptions of the HBM is that the beliefs of the individual
determine behaviour to a greater extent than the objective environment (Rosenstock,
1974). Salazar (1991) suggests that the model concentrates on phenomenological
aspects of the individual, to a lesser degree, the history or past experiences of the
person. The HBM proposes that the likelihood of a person adopting a given health-
related behaviour is a function of that individual's perception of a threat to their
personal health, and their belief that the recommended behaviour will reduce this
threat (Rosenstock, 1974). The key variables of the HBM include:
• Perceived threat:
Perceived susceptibility (an individual’s subjective perception of
the risk of contracting an illness/condition); and
Perceived severity (concerning the seriousness of contracting the
illness/condition);
• Perceived benefits (the believed effectiveness of strategies designed to reduce
the threat of the illness/condition);
• Perceived barriers (the potential negative consequences that may results from
taking particular health actions, including physical, psychological, and
financial demands);
• Cues to action (events such as physical or environmental, that motivate an
individual to take action);
• Self-efficacy (the belief in being able to successfully execute the behaviour
required to produce the desired outcomes) – introduced by Bandura in 1977;
and
• Other variables – diverse demographic, sociopsychological, and structural
variables that affect an individual’s perceptions and thus indirectly influence
health-related behaviour.
(Rosenstock, 1994)
373
Individual Perceptions Modifying Factors Likelihood of Action
Health Belief Model
Application to level crossing behaviour
In road safety, the HBM has been used in predicting and understanding safety
helmet use among schoolboy cyclists by analyzing attitudes and beliefs (Rutter, 1998)
as well as seat belt use among passengers (Simsekoglu, 2004). Driver behaviour at
level crossings could be explained by the health belief model by using the following
approach:
Likelihood of
behavioural change
Perceived threat of
illness/condition
Cues to action
Knowledge
Demographics
Perceived benefits versus barriers to
behavioural change
Perceived susceptibility
and seriousness of illness/condition
374
Concept Definition Application Perceived susceptibility
One’s opinion of chances of a collision
Define road user groups at risk; Personalize risk based on behaviours and features of groups; Heighten perceived susceptibility if too low
Perceived severity
One’s opinion of how serious a collision would be
Specify consequences of the risk and severity of a collision
Perceived benefits
One’s opinion of the efficacy of the advised action to reduce risk or seriousness of impact
Define action to take: how, where, when; clarify the positive effects to be expected
Perceived barriers
One’s opinion of the tangible and psychological costs of the advised action
Identify and reduce barriers through reassurance, incentives and assistance
Cues to action Strategies to activate “readiness”
Provide ‘how-to information’, promote awareness
Self-efficacy Confidence in one’s ability to take action
Provide training and guidance in performing action
Threat perception includes two components: susceptibility to a crash and
anticipated severity of the consequences of the crash (e.g. likelihood of being killed).
Behavioural evaluation consists of two distinct sets of beliefs: those related to barriers
(e.g. inconvenience for waiting for the train to pass, or peer pressure such as ‘beat the
train’) to safe level crossing behaviour and those concerning benefits (e.g. increased
safety). In addition to threat perception and behavioural evaluation, ‘cues to action’
and ‘health motivation’ components are included in the HBM. Cues to action refers
to triggers to safe level crossing behaviour (e.g. Signs alerting driver to danger ahead)
whereas health motivation refers to ones readiness to be concerned about health and
safety matters generally.
Criticisms of the health belief model
The health belief model has been criticized widely in the literature for a number
of reasons. Firstly, it has been criticized for not taking into account factors that are
involved in the decision-making process when performing a particular behaviour.
As the model is a psychosocial one, it can only account for variance in
behaviours which can be explained by attitudes and beliefs…Therefore, several
375
factors involved in the decision-making process (such as environmental,
intrapersonal and cultural factors) are neither measured nor identified.
(Toepell, 2003, p38)
Secondly, the prediction of behavioural intention (as opposed to the prediction
of actual behaviour) has been reported by Toepell (2003) to be a function lacking in
the model. Although not a criticism of the model as such, Toepell (2003) suggests that
very few studies implement the model in its entirety when predicting health
behaviour, and typically only use the variables of perceived susceptibility, severity,
barriers and benefits.
DETERRENCE THEORY
Overview
Deterrence theory has primarily been found in the domain of criminology and
sociology, although it has strong links to psychology. This theory has been used
extensively in Australia and other countries to guide the development and evaluation
of many road safety countermeasures (Homel, 1986, Elliott, 2003b, Watson, 2004a).
David South, an Australian psychologist who specialized in alcohol and traffic safety,
argues that “the reduction in the road toll…has arguably been the most successful
example of public action to minimize a social problem in Australia, and there is solid
evidence that general deterrence programs have played a major role” (1998, p76).
However, some road safety researchers would disagree. Harrison (1998) questions
South’s contention that there is ‘solid’ evidence supporting the role of general
deterrence, but recognizes that deterrence principles have played a pre-eminent role in
policy making for road safety in Australia.
Erickson, Gibbs and Jensen (1977) propose that deterrence theory should not
only be concerned with the objective properties of legal sanctions, but perceptions of
sanctions. Therefore, deterrence theory is primarily concerned with the manner in
which legal sanctions deter illegal behaviours through the mechanisms of fear and
perceived risk (Homel, 1986, Homel, 1988). Classical deterrence theory asserts that
the effectiveness of a legal threat is a function of the perceived certainty, severity and
376
swiftness of punishment (Homel, 1986, Vingilis, 1990). An important distinction is
made in deterrence theory between specific and general deterrence (Homel, 1986,
Akers, 1994). Specific deterrence relates to the individual level in which an offender
is deterred from re-offending through direct exposure to sanctions, while general
deterrence is at the population level in which the community is deterred through the
threat of legal sanctions (Homel, 1986). Stafford and Warr (1993) however, have
criticized classical deterrence theory for its failure to effectively account for the effect
of punishment avoidance on behaviour. They suggest that it is imperative to not only
consider the effect of a person’s direct experience with punishment and punishment
avoidance, but also their indirect or vicarious experiences obtained through interaction
with their peers (Stafford, 1993). As a consequence, Stafford and Warr (Stafford,
1993) proposed a reconceptualisation of deterrence theory to incorporate these key
principles. As a result of this reconceptualisation, specific and general deterrence are
no longer mutually exclusive processes operating on different populations, but can
operate conjointly on individuals (Watson, 2004a). This reconceptualisation of
deterrence theory has been supported by Piquero and Paternoster’s (1998) study in
examining drink driving behaviour. This study found that intentions to drink and
drive were affected by both personal and vicarious experiences, as well as experience
of punishment and punishment avoidance.
Application to level crossing behaviour
Enforcing traffic laws and deterring drivers from engaging in driving behaviour
that increases crash risk, has been found to be an effective way of reducing crash risk
in respect to drink driving and speeding behaviour (Cavallo and Cameron, 1992,
Cameron et al., 1992). According to Christie (2001), targeted deterrence and
enforcement measures have a high probability of changing driver behaviour, although
these measures are seldom popular with drivers and some sections of the media. This
efficacy of enforcement and deterrence is thought to be due to the higher risk of
detection for such behaviour relative to the risk of being involved in a collision
(Christie, 2001). According to Elliott (1992), this provides enforcement with a greater
potential to influence the motivation and behaviour of drivers.
In the arena of level crossing safety, enforcement is a mechanism that has
typically been overlooked. Red light cameras as a means of collecting data about road
377
user violations with possible use as a deterrent or enforcement mechanism have not
been used widely in Australia. However, the Australian Transport Council’s national
strategic direction to improve level crossing safety has highlighted the need for
ensuring legislation and enforcement are appropriate for the potential consequences of
a vehicle-train collision (Australian Transport Council, 2003). According to
deterrence theory, the decision to intentionally drive unsafely at level crossings (e.g.
drive around boom gates, drive through red flashing lights), should be mainly
influenced by a driver’s perceptions of the risk of apprehension, as well as the
certainty, swiftness and severity of punishment. Given the pre-eminent role of this
theory in road safety policy in Australia, it is very surprising that deterrence theory
has not been used to explain level crossing driver behaviour.
Criticisms of deterrence theory
Deterrence theory has been criticized by many experts for a number of reasons.
Vingilis (1990) has argued that deterrence theory fails to account for a wide range of
non-legal factors that may influence compliance with the law. These factors include:
social sanctions and rewards, moral commitment to the law, and the opportunity for
the commission of crime (Vingilis, 1990). Likewise, Akers (1990) argues that
deterrence theory is not a general or complete model of criminal behaviour as “the
primary concepts and valid postulates of deterrence and rational choice are
subsumable under general social learning or behavioural principles” (p655). Zimring
& Hawkins (1976) suggest that deterrence is aimed mostly at the marginal group who
lie between the law abiding citizens for who deterrence is not necessary and the
undeterrable citizens for who deterrence is ineffective. Ezzat Fattah, a leading
Canadian criminology academic, questions the real role of deterrence by suggesting
that its primary or direct effect may hide its real effects (Fattah, 1976). He suggests
that deterrence is less effective in controlling or eliminating habitual, unthinking
behaviour.
One would wonder whether the main function of legal sanctions with regard to
drivers’ behaviour is intimidation (curbing or modifying the behaviour through
the fear of punishment) or education (by instilling habitual modes of
performance in compliance with the traffic standards and regulations).
(Fattah, 1976, p48)
378
The conditions for which deterrence is more effective have been proposed by
Barry Elliott, a leading Australian road safety researcher. These conditions include:
• In controlling rational behaviour than in inhibiting impulsive behaviour;
• When motivation to engage in the prohibited behaviour is low and less
effective when the motivation is high (as in compulsive behaviour);
• When the behaviour is not habitual and requires thought processes and is
intentional – to the extent behaviour is already habitual, deterrence is least
effective;
• In assisting in the development of habitual modes of behaviour; and
• In controlling intentional behaviour (violations) rather than in eliminating
negligent behaviour (such as errors).
(Elliott, 2003a, p5-6)
SOCIAL COGNITIVE THEORY
Origins and overview
Bandura (1977b) first proposed the social learning theory (recently re-labeled
social cognitive theory) and asserted that the processes surrounding conforming
behaviour and deviant behaviour were similar. A basic tenet of this theory is that
social interaction informs and reinforces behaviour in which learning occurs through
direct experience and continues in ways that are dependent upon the negative and
positive consequences of that experience (Akers, 1979). Added to this tenet, is the
element taken from Bandura’s (1977b) social cognitive concepts, by which behaviour
can be learned through observation or modeling (Akers, 1990). Whilst deterrence
theory is concerned with the influence of legal sanctions on criminal behaviour, social
cognitive theory focuses on the overall social setting in which behaviours are
performed and the way in which they are differentially rewarded or punished (Akers,
1990).
Of the social cognitive theories that have been developed over time, Akers’ is
the most widely used to investigate a range of deviant or non-conforming behaviour.
Ronald Akers, in collaboration with Robert Burgess, developed a form of social
learning theory during the mid-1960’s, to explain criminal or deviant behaviour
379
(Akers, 1990, Akers, 1994). This theory drew on aspects of both sociological theory
(Sutherland’s Differential Association theory) and psychological theory (Skinner’s
operant conditioning and Bandura’s social learning concepts). Although this theory
has been developed over time, the fundamental concepts remain consistent with its
origins. This theory consists of four main concepts: differential association, imitation,
definitions and rewards. Differential association relates to the interaction and
subsequent identification with a group or an individual, with a person’s behaviour
being inclined to be compatible with that of the group. The most influential groups are
the family and peer group (Akers, 1979). The dimension of differential association
with this theory begins from interactions with the group with which one associates,
while the normative dimensions are those relating to the group’s evaluation of
different behaviours (Akers, 1996).
Imitation refers to behaviour that occurs after direct observation of that
behaviour or one similar by another person, with family and peer groups providing the
basis for most imitation or modeling (Akers, 1979). The media may be a powerful
source of imitation or modeling. Definitions are the norms, attitudes and orientations
that are learned through interaction with significant groups, with behaviour being
deemed as good or bad through these interactions and is learned by imitating the
behaviour of others (Akers, 1979). Verbal or cognitive definitions can be reinforced,
which occurs when the approval for the behaviour exceeds condemnation (Akers,
1979, Akers and Lee, 1996). Rewards (differential reinforcement) is the mechanism
through which behaviour is secured and sustained, and will either be strengthened or
weakened by reward or punishment (Akers et al., 1979). Akers’ (1979) proposes that
similar acts that are rewarded will provide rewards when one act is rewarded to a
greater extent than an alternative. Although Akers’ theory has demonstrated extensive
support in studies on both adolescent and adult populations, this theory has not been
utilized widely in the field of road safety.
Application to level crossing behaviour
In road safety, Aker’s social cognitive theory has been used to examine
unlicensed driving (Watson, 2004b) as well as analyzing road accidents involving
child pedestrians (Howarth and Repetto-Wright, 1978, Howarth, 1974). Watson’s
(2004b) study found that Akers’ theory was better able to predict unlicensed driving
380
than deterrence theory, which was thought to be due to the more comprehensive
framework that social cognitive theory provides over deterrence theory. Additionally,
social cognitive theory was able to include the contributions that the attitudes of
family, friends and others have in developing personal attitudes towards behaviour
(Watson, 2004b).
Although social cognitive theory has not been applied to level crossing safety to
date, it may offer some heuristic advantages over other theories. Firstly, it provides a
comprehensive method of addressing a variety of important factors such as formal and
informal sanctions, direct and indirect experiences, and punishment and punishment
avoidance. Secondly, in applying this theory to level crossing behaviour, the primary
process that influences unsafe behaviour would be through an individual’s interaction
and involvement with their social groups (differential associations). Involvement in
such social groups allows opportunities for the imitation of models and learning of
attitudes (either positive or negative) towards behaviours such as risk taking at level
crossings. The likelihood of unsafe driving at level crossings would increase with
either neutral or favorable definitions towards the behaviour. The individual’s
anticipated balance of reinforcement for unsafe driving behaviour at level crossings
(differential reinforcement), further contributes to a decision to take risks or not. This
is dependent upon an individual’s prior and current rewards and punishments for risk
taking at level crossings, as well as anticipated rewards and punishments for different
behaviour.
Criticisms of social cognitive theory
Although this theory has been used widely in a variety of settings, Lapinski and
Witte (1998) suggest that one element missing from the model is motivation.
According to Delaney, Lough, Whelan and Cameron (2004), the theory appears to
assume that people with high levels of self-efficacy and outcome expectations
consistent with the recommended response, would be motivated to act. Other
researchers have reported that normative pressures have little impact on health
behaviour (deVries, 1990, Kok, 1991, Lechner, 1995). According to Bandura (1998)
this raises the question of whether normative influences are ineffectual or whether
they need to be measured more comprehensively as different forms of social outcome
expectations.
381
SENSATION SEEKING
The possibility that sensation seeking underlies risk taking in driving has been
the subject of a considerable amount of research during the past few decades. The
interest in driver’s personality as a potential underlying causal factor in driver
behaviour was brought about by Tillman and Hobbs (1949) proclamation that ‘a man
drives as he lives’ (Jonah, 1997). Accordingly, sensation seeking has been found to
be higher in males than females (Jonah, 1997). Age is another factor that is relevant
for the motivation to seek sensation. Studies indicate that there is a decline in
sensation seeking with age (Giambra et al., 1992), with one study showing that it rises
between ages 9 and 14, peaks around age 20 and then declines steadily thereafter
(Zuckerman, 1994b).
Studies have demonstrated that there is a significant positive relationship
between risky driving and the theoretical construct of sensation seeking (Jonah, 1997).
Jonah’s (1997) literature review of this relationship found that the majority of the 40
studies reviewed showed a positive relationship between sensation seeking and risky
driving, with correlations in the 0.3-0.4 range. These studies focused mainly on drink
driving, speeding and following too closely. Additionally, these studies have found a
stronger relationship between self-reported driving behaviour and sensation seeking,
rather than with crash involvement or traffic offences, which may be a result of
weaknesses in crash measures used. This review postulates that high sensation seekers
adapt their behaviour to take advantage of safety improvements (e.g. protection
systems) to a greater extent than low sensation seekers, and that consequently they
exhibit greater risky driving (Jonah, 1997). Some researchers believe that sensation
seeking tendencies can not be controlled but rather a trait.
According to Zuckerman (1994a) sensation seeking “is a trait defined by the
seeking of varied, novel, complex, and intense sensations and experiences and the
willingness to take physical, social, legal and financial risks for the sake of such
experiences” (p27). Central to this trait is the optimistic tendency to approach novel
stimuli and explore the environment (Jonah, 2001). The broader trait referred to by
Zuckerman (1994a) as ‘impulsive sensation seeking’ is closely associated with
Eysenck’s (1983) ‘psychoticism dimension’. Evidence suggests that individual
differences are embedded in biological differences in the brain and may in fact to
some extent be hereditary (Zuckerman, 1994b, Eysenck, 1983). Zuckerman (1994b)
382
proposes that monamine neurotransmitters like dopamine, norepinephrine and
serotonin underlie the trait of sensation seeking. Zuckerman developed a
psychometric test to diagnose individual differences based on the bio-developmental
sources of an individual’s motivation to sensation seek (Eysenck, 1983, Zuckerman,
1994b, Horvath and Zuckerman, 1993). This ‘sensation seeking score’ is one measure
in terms of how sensation seeking is operationally defined. Although Zuckerman’s
test has been subjected to empirical validation (Burns and Wilde, 1995), it has been
found that there is a relatively moderate empirical connection between the test and
real-life reflections of the related disposition (Rosenbloom and Wolf, 2002).
Application to level crossing behaviour
According to Witte and Donohue (2000), the lack of perceived susceptibility
toward vehicle-train collision may be due to either errors in judgment, sensation
seeking tendencies or a lack of personal experience with someone who has been
involved in a collision at a level crossing. Jonah (1997) proposes that:
Some high SSs [sensation seekers] will not be influenced by educational
measures since these drivers actually enjoy the thrill of risky driving and
avoiding negative consequences. Furthermore, enforcement programs may fail
to deter some high SSs’ risky driving since they enjoy the thrill of breaking the
law and avoiding detection. These drivers may only be protected by engineering
solutions, whereby they do not have to alter their behaviour to improve their
safety (e.g. airbags, anti-lock brakes). On the other hand, it is conceivable that
high SSs would readily adapt to any perceived improvements in their level of
risk by accepting greater risks.
(p663)
Results from Witte and Donohue’s (2000) formative evaluation suggest that the
majority of respondents engage in safe driving at level crossings, with 10-20%
reporting extremely risky behaviours such as trying to ‘beat the train’. The authors
suggest that high sensation seekers are motivated to avoid frustrating and dull
experiences, admitting they have increased susceptibility to being in a vehicle-train
collision (Witte, 2000). These findings also suggests that risky drivers are habituated
383
to high levels of fear because of prior close calls with trains, and engage in biased
judgments about their abilities to ‘beat the train’ because they have ‘made it’ before.
It is difficult to discern as to what role sensation seeking plays in fatal collision
at level crossings. Previous research conducted by Wigglesworth (1979) suggests that
sensation seeking has not played a role in fatal level crossings collisions. This study of
85 consecutive fatal crashes involving motor vehicles at all types of level crossings in
Victoria, examined the detailed police reports prepared for the coroner. Wigglesworth
(2001) suggests that these reports indicate that “in most cases, the accident occurred
to a law-abiding citizen going about his or her daily work and was attributable to
human overload unrelated to any breach of the regulation” (p311). Since this time,
some may argue that sensation seeking of drivers has increased. However, due to
limited data supporting the role of sensation seeking in fatal crashes, this construct
was not included in any data collection measures.
HABITUAL BEHAVIOUR
According to Ouellette and Wood (1998), habits can be defined as “behavioural
tendencies to repeat responses given a stable supporting context” (p55). Such
repetition and practice of a response in a given setting, allows cognitive processes that
initiate and control the response to become automatic and be performed quickly in
parallel with other activities and little focal attention (Goldenbeld, 2000). Sutton
(1993) argues that almost all behaviours are able to be repeated and that many health
behaviours must be repeated if they are to be effective. For an individual, habits are
essential in handling every day activities as they facilitate information processing
contributing to the efficiency of behavioural routines (Goldenbeld, 2000). Older
people develop individual lifestyles in which habits play a crucial role.
Repeated behaviours however, create an important problem for the theories of
reasoned action and planned behaviour (Sutton, 1993, Ronis, 1989). Although the
theory of planned behaviour incorporates automatic processes, it generally assumes
reasoned processes underlying attitudes and actions (Ajzen, 2002b). This theory
postulates that human social behaviour is reasonable (Bamberg, 2003), however many
theorists challenge this standpoint and argue than human behaviour can be automatic
or habitual (Aarts et al., 1998, Aarts and Dijksterhuis, 2000, Ouellette, 1998, Triandis,
384
1977, Ronis, 1989, Fazio, 1990, Bagozzi, 1981). It is proposed that when behavior is
performed repeatedly and becomes habitual, it is guided by automated cognitive
processes, rather than being preceded by complex decision processes (i.e. a decision
based on attitudes and intentions) (Aarts et al., 1998). Hence, it is based on habitual
or automatic processes, rather than reasoned decision-making. To test this assumption,
past behaviour measures play an important role. It is argued that if social behaviour
is reasoned, frequency of prior behaviour should have only an indirect link to later
behaviour with its effect being mediated by intention and perceived behaviour control
(Bamberg, 2003). Bamberg and colleagues (2003) point out that if added to the
regression equation, past behaviour is normally found to significantly improve the
prediction of later behaviour over and above the effects of intentions and perceptions
of behaviour control. It can be said that, such findings indicate that rather than being
completely reasoned, the behaviour is partly under the direct control of the stimulus
situation (i.e. that it habituates with repeated performance) (Bamberg, 2003). This
perspective considers that frequency of past behaviour is an indicator of habit
strength, which can be used as an independent predictor of later action (Bamberg,
2003).
According to Ajzen (2002b) strong and unmediated links between prior and
later behaviour imply habituation in a process that bypasses intentions. This is often
taken as evidence for habituation of behaviour and as complementing the reasoned
mode of operation assumed by behaviour-intention models such as the theory of
planned behaviour (Ajzen, 2002b). When a habit starts to develop, behaviour is
thought to come under the control of stimulus cues and on future occasions, presence
in a similar situation is sufficient to activate the automatic response chain (Ajzen,
2002b, Aarts et al., 1998, Ouellette, 1998). Therefore, for habitual behaviour to
occur, a stable context is crucial. Thus, habit is defined as the tendency to repeat past
behaviour in a stable context (Ouellette, 1998).
‘Routinization of behaviour’ is also consistent with a reasoned action
perspective according to Ajzen (2002b). The habituation viewpoint affirms that
routinized behaviour is under the control of stimulus cues, however, the reasoned
action perspective proposes that such behaviour is guided by automatically activated
or spontaneous attitudes and intentions (Ajzen, 2002b). According to the principles of
the theory of planned behaviour, attitudes and intentions – once formed and well-
established – are activated automatically and guide behaviour without the requirement
385
of conscious supervision (Ajzen and Fishbein, 2000). Thus, it does not propose that
individuals review their behavioural, normative and control briefs prior to every
enactment of a frequently performed behaviour (Ajzen, 2002b).
Road safety researchers have argued that to a large extent, driver behaviour is
influenced by habit. Seat belt usage (Bentler, 1979) and travel mode choice are two
behaviours that are claimed to be habit-like behaviours. Sutton et al. (1990) in their
study found that there is a substantial habitual component to wearing seat belts and
that this behaviour is not under volitional control. Wittenbraker et al. (1983) found
similar results in their study and argue that prior behaviour, or habit should be added
to the model of reasoned behaviour. More recently, Hatakka, Keskinen, Katila and
Laapotti (1997) found in their study that self-reported risk driving habits and skills for
careful driving, were good predictors for traffic violations and crashes in both
correlational and longitudinal methodologies.
Application to level crossing behaviour
With regards to level crossing safety, driving over the same level crossing/s on a
regular basis eventually becomes a habit in the sense that little cognitive effort is
required for continued execution of the behaviour. So long as the context remains
relatively unchanged, such routinized behaviour is performed in a largely automatic
manner with little conscious control. Research conducted by Wigglesworth (1979)
indicates that of 85 consecutive fatal crashes in Victoria during the 1970’s, 86% of
those drivers killed in a level crossings collision were local residents who were
familiar with the crossing. However, this research does not detail how frequently the
driver was exposed to driving at this level crossing.
The perspective argued by Bamberg and colleagues (2003) that past behaviour
is an indicator of habit strength, may provide useful to examining habitual behaviour
at level crossings in the current study. Although it is not possible to test for
significant associations between past behaviour and collisions at level crossings in the
current study, an instrument pertaining to self-report past behaviour will be included
to test for any association between past behaviour and involvement in any crashes
during the past year.
386
Appendix 2: Modified Delphi Technique (First Questionnaire)
387
Dear Panelists,
Thank-you for agreeing to take part in this 1st questionnaire on Motorist
Behaviour at Railway Level Crossings. You have been identified as a key stakeholder
and we are seeking your participation in this project. There will be 2-3 short survey
questionnaires to complete during this project.
This 1st questionnaire asks you to write your answers in the space provided after
each question. We will then combine all the answers to give us a broad overview of
all the issues. The 2nd questionnaire which will be sent to you in late August, will
give you a summary of the findings from this 1st questionnaire and then you will be
asked to rank issues/items for importance. Depending on whether we require more
information, a 3rd questionnaire will be sent to you. This 3rd questionnaire will
summarize the 2nd questionnaire and also ask you to rank issues/items of importance.
Your individual contribution to each questionnaire is very important to
improving level crossing safety in Australia. If you would like to clarify any question
in the 1st questionnaire, please contact the Project Officer (Ms Nadja Ibrahim) on 07
3864 4926 or email: [email protected]
Dr. Jeremy Davey
Chief Investigator
Centre for Accident Research & Road Safety - Qld
388
Questions:
• What do you believe are the major factors contributing to vehicle-train
collisions at railway level crossings?
• Please comment on the current engineering devices/design at level crossings
and their impact on the safety of motorists.
• What types of motorists or vehicles do you consider are most at risk of being
involved in a crash?
• In your opinion, what specific things do motorists do to contribute to vehicle-
train collisions at level crossings? i.e. What are the behaviours that put them
at risk?
• Do you think motorists behave differently between active (lights and/or boom
gates) and passive crossings (no protection systems)? If yes, how do they
behave differently?
• In your opinion, why do motorists behave in this way at level crossings?
• In your opinion, do interactions between different motorists and/or vehicles
contribute to vehicle-train collisions?
389
Demographics
What State/Territory do you currently work in? ______________
Which field do you work in?
Road Rail Other (Please Specify) ________________________
How long have you worked in this current field?
_______________ yrs
Which type of organisation do you mainly work for? Please choose one only.
National Government Organisation State Government Organisation National Non-Government Organisation State Non-Government Organisation Police Service Academic Institution Private Business/Company Other (Please Specify) ________________________
Which of these areas within your organisation are you mainly involved in? Please choose one only.
Policy Project Management Research and/or Development Enforcement Engineering Education Communications/Marketing Other (Please Specify) ________________________
What is your gender?
Male Female
390
Appendix 3: Modified Delphi Technique (Second Questionnaire)
391
Dear Panelists,
We are now seeking your participation in the next round of these questionnaires
on Motorist Behaviour at Railway Level Crossings.
The Centre for Accident Research and Road Safety – Queensland (CARRS-Q)
based at the Queensland University of Technology, is investigating motorist
behaviour at level crossings. The first phase of this 3-year project is to identify (by
key stakeholders) target behaviours and target groups that may place motorists at risk
of a vehicle-train collision.
Your opinion is very valuable to us and we thank-you for your time and
participation. If you would like any further information about this project, please do
not hesitate to contact me or the project team
Dr. Jeremy Davey
Chief Investigator
Centre for Accident Research & Road Safety - Qld
392
Major Contributing Factors
A number of factors contributing to vehicle-train collisions at railway level crossings were identified in the previous questionnaire. Please indicate how important you believe each of these is to a collision or a near miss. Circle one number on the scale for each factor, according to how important you believe the factor is in regards to vehicle-train collisions or near misses.
Ver
y im
port
ant
Not
at a
ll im
port
ant
Motorist Behaviour
1. Trying to beat the train across the crossing 1 2 3 4 5
2. Inattention by the motorist when driving 1 2 3 4 5
3. Disobeying the warning signs or signals at crossings 1 2 3 4 5
4. Speeding on approach to crossing 1 2 3 4 5
5. Queuing over crossings 1 2 3 4 5
6. Not stopping at ‘stop’ signs at passive crossings 1 2 3 4 5
7. Not slowing down to scan for a train at passive crossings 1 2 3 4 5
8. Driving around closed boom gates at active crossings 1 2 3 4 5
9. Alcohol / drugs use 1 2 3 4 5
10. Becoming confused at active crossings 1 2 3 4 5
Motorist Error
11. Misjudging train speed 1 2 3 4 5
12. Misjudging distance of train from the crossing 1 2 3 4 5
13. Misjudging the time needed to cross safely 1 2 3 4 5
14. Not seeing level crossing on approach 1 2 3 4 5
Motorist Expectations
15. Low expectation of coming across a train while driving 1 2 3 4 5
16. Familiarity with train timetables 1 2 3 4 5
17. Low perceived chance of a crash occurring 1 2 3 4 5
18. Perception of long delays when stopping for a train to pass 1 2 3 4 5
19. Not expecting a second train
1 2 3 4 5
Motorist Lack of Knowledge, Training or Experience
20. Urban motorists’ inexperience with passive level crossings
1 2 3 4 5
21. Lack of education / training on level crossing safety during 1 2 3 4 5
393
initial driver licensing stage
22. Lack of awareness about the consequences of driving unsafely at
level crossings (i.e. severity of crashes)
1 2 3 4 5
23. Lack of knowledge about a trains stopping distance 1 2 3 4 5
24. Poor knowledge of the road rules in relation to level crossings 1 2 3 4 5
Engineering Issues
25. Road markings are ineffective as they are usually ignored by
motorists
1 2 3 4 5
26. Inadequate sighting distances for motorists at crossings 1 2 3 4 5
27. Traffic congestion leading to queuing over crossing 1 2 3 4 5
28. Inadequate warning signage at crossings 1 2 3 4 5
29. Short stacking situations (i.e. where there is insufficient stacking
distance from the nearest road intersection for the crossing to be
cleared safely by long vehicles)
1 2 3 4 5
30. Inadequate warning devices at crossings 1 2 3 4 5
31. Environmental/design issues affecting ability to see warning
devices (e.g. sun glare)
1 2 3 4 5
32. Inconsistent signage between crossings 1 2 3 4 5
33. Inconsistent approaches by authorities to risk treatment /
management at crossings
1 2 3 4 5
34. Human visual limitations making speed judgment of
approaching trains difficult
1 2 3 4 5
Lack of Enforcement
35. Lack of enforcement of road rules at level crossings by police 1 2 3 4 5
36. Low fines and penalties for potential severity of collision 1 2 3 4 5
37. Lack of detection devices for breaches (no “red light” cameras at
level crossings)
1 2 3 4 5
394
Road User Groups
A number of road user groups at risk of vehicle-train collisions were identified in the previous questionnaire. These include: general motorists, younger motorists, older motorists, long haul/heavy vehicle drivers, rural motorists, bus drivers, local drivers and fleet drivers. For each of the road user groups, there are a number of factors that may contribute to their risk of a vehicle-train collision. Please indicate how important you believe each of these factors is to each group.
Ver
y im
port
ant
Not
at a
ll im
port
ant
General Motorists
1. Errors of judgment (e.g. misjudging time needed to cross safely) 1 2 3 4 5
2. Speeding on approach to crossing 1 2 3 4 5
3. Motorist fatigue 1 2 3 4 5
4. Trying to beat the train across the crossing 1 2 3 4 5
5. Driving impatiently 1 2 3 4 5
6. Lack of knowledge about safe driving behaviour at crossings 1 2 3 4 5
7. Misjudging level of risk 1 2 3 4 5
8. Driving while distracted (e.g. talking on a mobile phone) 1 2 3 4 5
9. Queuing over crossing 1 2 3 4 5
10. Community norms relating to behaviour at level crossings (e.g.
“no-one stops at this crossing”)
1 2 3 4 5
Younger Motorists (17-24 years)
11. Trying to beat the train across the crossing 1 2 3 4 5
12. Inexperience with crossings 1 2 3 4 5
13. Driving impatiently 1 2 3 4 5
14. Misjudging level of risk 1 2 3 4 5
15. Lack of training on crossings during licensing process 1 2 3 4 5
Older Motorists (60 years and over)
16. Vision impairment 1 2 3 4 5
17. Confusion/anxiety at crossings 1 2 3 4 5
18. Slow reaction times 1 2 3 4 5
19. Decision making difficulties 1 2 3 4 5
20. Errors of judgment (e.g. misjudging time needed to cross safely) 1 2 3 4 5
395
Long Haul/Heavy Vehicle Drivers
21. Difficulties with vehicle braking and acceleration 1 2 3 4 5
22. Trying to beat the train across the crossing 1 2 3 4 5
23. Length of vehicle causing overhang on the crossing 1 2 3 4 5
24. Risk taking due to commercial time pressures 1 2 3 4 5
25. Extended time taken to safely clear crossing 1 2 3 4 5
26. Truck driver fatigue 1 2 3 4 5
27. Low expectation of a train being on crossing 1 2 3 4 5
28. Not scanning for a train at ‘give way’ signed crossings 1 2 3 4 5
29. Trying to beat the train across the crossing 1 2 3 4 5
30. Complacency due to familiarity with crossings 1 2 3 4 5
31. Community norms relating to behaviour at level crossings (e.g.
“no-one stops at this crossing”)
1 2 3 4 5
Bus Drivers
32. Bus driver distraction (e.g. school bus driver) 1 2 3 4 5
33. Extended time taken to safely clear crossing 1 2 3 4 5
34. Trying to beat the train across the crossing 1 2 3 4 5
35. Difficulties with vehicle braking and acceleration 1 2 3 4 5
36. Queuing over crossing 1 2 3 4 5
37. Risk taking due to commercial time pressures 1 2 3 4 5
Local Drivers (motorists that live within 15km of a level crossing)
38. Low expectation of a train being on crossing 1 2 3 4 5
39. Queuing over crossing 1 2 3 4 5
40. Complacency due to familiarity with the crossings 1 2 3 4 5
41. Inattention while driving 1 2 3 4 5
42. Trying to beat the train across the crossing 1 2 3 4 5
43. Driving impatiently 1 2 3 4 5
44. Community norms relating to behaviour at level crossings (e.g.
“no-one stops at this crossing”)
1 2 3 4 5
Fleet Drivers
45. High level of exposure to crossings (i.e. drive over numerous
crossings per day)
1 2 3 4 5
46. Speeding on approach to crossing 1 2 3 4 5
47. Motorist fatigue 1 2 3 4 5
396
48. Lack of knowledge about safe driving behaviour at crossings 1 2 3 4 5
49. Misjudging level of risk 1 2 3 4 5
50. Risk taking due to commercial time pressures 1 2 3 4 5
51. Driving while distracted (e.g. talking on a mobile) 1 2 3 4 5
397
Appendix 4: Survey instrument used in Study Two
398
MOTORIST BEHAVIOURS AT RAILWAY LEVEL CROSSINGS
In our recent research with the general public, the following behaviours were reportedly performed at railway level crossings. Please indicate what level of risk you would attribute to each behaviour, in terms of risk of a
vehicle-train collision.
Please put one “X” in the appropriate box for each of the behaviours.
Low
ris
k
Som
e ri
sk
Mod
erat
e ri
sk
Hig
h ri
sk
Ver
y hi
gh r
isk
1. Driving through the crossing when the lights are flashing but before the boom gates start to go down
2. Trying to drive through the crossing when the boom gates are actually coming down
3. Driving through the crossing when the lights are flashing and the boom gates have started start to go up
4. Slowing and rolling through STOP signed crossings if no train visible (rather than coming to a complete stop)
5. Driving through the crossing as soon as one train has passed, without looking for a second train, at a passive or crossing with lights only
6. Not stopping at all at STOP signed passive crossings 7. Not looking for trains at passive crossings 8. Drive through passive crossings when the train is visible but still
some distance away
9. Drive through passive crossings when the train close to the crossing 10. Queuing over a congested crossing 11. Driving around the boom gates 12. Try to beat the train over the crossing 13. Speeding on approach to crossings 14. Overtaking cars that are stopped at either an active or passive
crossing
15. Looking or scanning on approach to a passive crossing and then not topping or slowing if no train seen
16. Driving through a crossing when the lights are flashing and the train is not visible (at crossings with lights only)
17. Driving through a crossing when the lights are flashing and the train is visible (at crossings with lights only)
18. At passive crossings, following the car in front across the crossing without looking
19. Driving across the crossing when unable to see if there is a train coming (poor visibility due to shrubs, cane, parked cars etc)
Thank you for your time and participation
399
Appendix 5: Support letter from RACQ inviting Younger Drivers to participate
400
29th November 2006 Dear Member
Younger Driver Safety Study The RACQ is calling for volunteers to be a part of a ground-breaking study to understand the driving behaviour of young drivers aged 17 - 24 years. As part of its ongoing commitment to road safety, the RACQ actively supports the valuable research work carried out by the Centre for Accident Research and Road Safety – Queensland (CARRS-Q) at Queensland University of Technology. CARRS-Q is a joint initiative of the Motor Accident Insurance Commission and QUT, which conducts applied research into areas of road safety and accident prevention as well as administering a road safety research grant scheme on behalf of the MAIC. As the RACQ’s representative on the CARRS-Q Board of Management, I am acutely aware of the significance of the work carried out by the Centre and its value to Queensland motorists. That is why I am pleased to draw your attention to this innovative research study into younger driver behaviour to be undertaken by the RACQ and CARRS-Q in order to develop more effective road safety campaigns. The RACQ is seeking your participation. Very little time is required of participants. Online surveys will take approximately 20 minutes to complete. In total, 2 surveys will need to be completed. As an incentive for young drivers to participate, 2 movie tickets will be posted to you on completion of the questionnaires. Participants will be asked a range of questions about road safety campaigns and their driving behaviour. Please be assured that any personal information provided by you will remain confidential and used solely for statistical research purposes. To participate in this research, you will need to access one of the websites below and use login details provided. Please note that to assist the online server in coping with internet traffic from the survey, two separate websites are being used. We ask you to read the following directions to ensure that you access the correct site. If you were born in the months January to June, use the following details to participate:
Website: http://rail.carrsq.net.au/S212/ Username: drive2006 Password: road06
If you were born in the months July to December, use the following details to participate:
Website: http://rail.carrsq.net.au/S313/ Username: motor2006 Password: road06
Thank you for your cooperation. Yours sincerely Gary Fites General Manager External Relations
401
Appendix 6: Support letter from RACQ inviting Older Drivers to participate
402
15 May 2006
Dear Member
Older Driver Safety Study The RACQ is calling for volunteers to be a part of a ground-breaking study to understand why drivers aged 60 years and over are involved in more road crashes. As part of its ongoing commitment to road safety, the RACQ actively supports the valuable research work carried out by the Centre for Accident Research and Road Safety – Queensland (CARRS-Q) at Queensland University of Technology. CARRS-Q is a joint initiative of the Motor Accident Insurance Commission and QUT, which conducts applied research into areas of road safety and accident prevention as well as administering a road safety research grant scheme on behalf of the MAIC. As the RACQ’s representative on the CARRS-Q Board of Management, I am acutely aware of the significance of the work carried out by the Centre and its value to Queensland motorists. That is why I am pleased to draw your attention to this innovative research study into older driver behaviour to be undertaken by the RACQ and CARRS-Q in order to develop more effective road safety campaigns. While age is not the sole indicator of driving ability, drivers aged 60 years and over are involved in more fatal crashes per kilometres travelled than other drivers. Gradual declines in visual, cognitive and motor functions can make drivers vulnerable to crashes in complex situations that require high levels of visual perception, attention, and rapid response. As a result, older drivers are more likely to be involved in crashes at intersections and in other such traffic situations. The RACQ is seeking your participation. Very little time is required of participants. Surveys will take approximately 20 minutes to complete and will be sent to you in the mail (with reply-paid envelopes included). In total, 2 surveys will need to be completed. If you would like to participate in this research, please phone Mr Nicholas Stenson on 07 3864 4926 or email [email protected] Survey questionnaires will be sent to you in the mail and you will then be phoned by a member of the research team to ask your opinion about road safety campaigns that you may have seen or heard. Please be assured that any personal information provided by you will remain confidential and used solely for statistical research purposes. For your participation, you will be sent 2 movie tickets once you have completed the surveys. Thank you for your cooperation. Yours sincerely Gary Fites General Manager External Relations
403
Appendix 7: Intervention and Control Radio Script for Each Road User Group
404
Heavy vehicle intervention (radio script)
Pre-amble
Accidents at railway crossings are like lighting strikes. They don’t happen all
the time and you never think it will happen to you, but, if it does, you probably won’t
survive it. Every year, though, trucks hit trains, lives are lost and millions of dollars
of damage is done.
Source introduction
If you’ve been driving trucks up and down Australia, you will know what
happens when a train hits a B-Double, and you’ll know it’s not pretty.
Suggested action
We’re not going to tell you how to do your job. If you’re a truck driver, you
know the roads and you know where the dangerous spots are. Use your experience to
protect yourself at railway crossings:
• Think about Stopping Distances – some crossings have lights and boom gates,
others just have a stop or give-way sign, but all require you to stop. Prepare
early, make sure you give your truck enough time to pull up.
• Look out for Short-stacking – this is when your truck hangs over the crossing
because there is not enough space ahead to clear it. Make sure that your path
is clear before you start to cross. You don’t want to realise your stuck half-
way.
• Check for Problems with Visibility - some conditions can make it harder to
check for trains at crossings. Don’t gamble at s-bend roads, in glare or when
roadside objects obscure your view. Slow down and make sure you are safe to
cross.
Re-cap
Work is currently underway fixing up the roads at level crossings, but with
thousands of railway crossings in Australia, it is going to take some time to fix them
405
all up. Until then, think about the crossings you pass, and take care to protect yourself
and others. So remember…Keep your brain in gear: Stay alert at railway crossings.
Official end
This message is brought to you by the Motor Accident Insurance Commission
in the interests of road safety.
406
Older Driver Intervention (radio script)
Pre-amble
Each year, accidents at railway level crossings claim the lives of Australian
drivers. It’s tempting to think that these deaths are hoons or inexperienced drivers.
But often those killed at level crossings are just normal people. Twenty-six percent
(26%) of them are aged over 60.
Source introduction
As older drivers, we have a wealth of experience behind the wheel, but this
doesn’t make us immune to the dangers on the road. While we are very conscious of
day to day risks, sometimes it can be the unexpected danger, the once in a lifetime
event, that we are not prepared for.
Suggested action
This is what you need to know to protect yourself at level crossings:
• Know your crossings – some have lights and boom gates, others just have a
stop or give-way sign, but all require you to give way to trains.
• Plan your trip – know what crossings you are likely to pass and how to
negotiate them. If you find certain crossings confusing, ask someone you trust
for advice.
• Take your time – no one has the right to rush you on the road. Following the
rules at level crossings makes the road safer not only for yourself, but for other
drivers as well.
Re-cap
The stakes at railway crossings are high. Being prepared may not just save your
life, but those of the people you care most about. So remember, every time you use a
level crossing…Be Cautious. Be Smart. Be Safe.
407
Official end
This message is brought to you by the Motor Accident Insurance Commission
in the interests of road safety.
408
Older Driver Control (radio script)
Pre-amble
Your eyes are one of the most important senses for driving. As drivers get
older, their eyesight is less clear and they cannot see as far.
Source introduction
For drivers aged 60 years and over, it is often hard to see detail such as traffic
signs, and it may be difficult to see objects approaching from the side. Handling glare
and recovering from glare is another problem for mature drivers. As such, mature
drivers can have problems with oncoming headlights or the afternoon sun. Eyesight is
also not as good at twilight or at night time.
Suggested action
If you are aged 60 years and over, it is important to maintain your eyesight to
drive safely. Two things you can do to make sure your eyesight is at its best are:
• Have regular check-ups at the Optometrist at least every two years and
• Keep the prescription of your glasses up-to-date
Other things you can do to ensure you drive safely are always to make sure your
windscreen is clean to reduce glare and limit your driving at night or at twilight.
Re-cap
So next time you walk past an Optometrist, book in for a check-up. It may just
save your life.
Official end
This message is brought to you by the Motor Accident Insurance Commission
in the interests of road safety.
409
Younger Driver Intervention (radio script)
Pre-amble
Each year, accidents at rail level-crossings claim the lives of young Australian
drivers. Some of them are taking stupid risks. Some weren’t paying attention. But
most of them get hit just trying to save a little bit of time.
Source introduction
When you see what happens when a train hits a car, you don’t ever forget the
carnage, the twisted metal and the lives lost. What you don’t often see is the long-
term damage they do: to families, to mates and to lives of those left behind.
Suggested action
When it comes to accidents at level-crossings, there aren’t any second chances.
This is what you need to know to protect yourself at these crossings:
• Know your crossings – some have lights and boom gates, others just have a
stop or give-way sign. It’s important to remember, however, that all of these
crossings require you to stop.
• Know your trains – trains travel at speeds up to 100 km/hr and way over 120
tonnes. At that speed and with size, they can take up to a kilometre to stop.
Don’t assume you can beat them across the tracks, and don’t think they will be
able to stop in time to avoid a collision if you try to.
• Know your facts – the stop time at most level-crossings is shorter than you
think, around the same time as a change of traffic lights. In the long-run,
adding an extra minute to your journey makes no difference to your arrival
time, but it might save your life.
Re-cap
So remember, every time you reach a level-crossing, think about what you’re
doing. A minute of your time can save a lifetime of regrets.
410
Official end
This message is brought to you by the Motor Accident Insurance Commission
in the interests of road safety.
411
Younger Driver Control (radio script)
Pre-amble
Road safety education has been talking about seatbelts for a long time.
However, statistics show us that small group of car occupants, both drivers and
passengers, still routinely travel without seatbelts. If that’s not enough, the hospital
wards are full of proof that some people still aren’t getting the message.
Source introduction
So this time, we’ll let the numbers speak for themselves. 50 – the number of
Queenslanders killed each year in car accidents who weren’t wearing seatbelts. 244 –
the number of Queenslanders hospitalised each year because of car accidents, who
weren’t wearing seatbelts. 10 – the number of times more likely you are to die in a
car accident, if you aren’t wearing a seatbelt.
Suggested action
The numbers don’t lie. A seatbelt can be the difference between walking away
from an accident, and never walking again.
Re-cap
Next time you get into a car, take a second and buckle up. It’s that simple.
Official end
This message is brought to you by the Motor Accident Insurance Commission
in the interests of road safety.
412
Appendix 8: Questionnaires used in Study 3
413
ID Number: __ __ __ - __ Staff Use Only
Centre for Accident Research and Road Safety – Queensland
Young Drivers Pre-Test Questionnaire
2006
Associate Professor Jeremy Davey Deputy Director Centre for Accident Research & Road Safety – Queensland Queensland University of Technology Telephone: 07 3864 4574 Email: [email protected]
Angela Wallace PhD Scholar Centre for Accident Research & Road Safety – Queensland Queensland University of Technology Telephone: 0402 240 234 Email: [email protected]
414
1. Personal Information We would like to start by asking you some questions about yourself. Please circle the number that best describes you.
What is your gender? Male………………………………………………….……
Female…………………………………………………… 1 2
What is your age? _________ Years
Are you currently involved in any paid work?
Yes, I am currently working …………………………… No, I am not currently working ………………………..
1 2
If you are in paid work, what is your occupation?
_______________________________________
What is the postcode where you live?
What type of drivers licence do you hold? Learners ……....………………………………………… Provisional ……………………………………………… Open licence …………………………………………….
1 2 3
2. Education What is the highest level of education that you have completed? Please circle one number that best describes your answer. Some secondary or high school …………………………………………………………………………….. Year 12 …………………………………………………………………………………………………………. Some university …………………………………………………………………………………………………. University degree ………………………………………………………………………………………………... Some TAFE ……………………………………………………………………………………………………… TAFE Diploma or Certificate …………………………………………………………………………………… Other (Please specify) ______________________________________________________________
1 2 3 4 5 6 7
3. Driving Patterns We would like to ask you some questions about how often you drive.
On average, how many days of the week do you drive? ______________________ days per week
On average, how many hours per day do you drive? ______________________ hours per day
415
4. Driving Self-Assessment
The next few questions relate to your driving. We would like for you to answer these questions as truthfully as possible. (Circle one number for each statement) How often do you…..
Nev
er
All
the
time
Wear your seatbelt while driving 0 1 2 3 4 5
Drive more than 10km/hr above the speed limit in built-up areas 0 1 2 3 4 5
Stop at ‘Stop’ signs 0 1 2 3 4 5
Feel tired when driving at night-time 0 1 2 3 4 5
Drive more than 10km/hr above the speed limit on open roads 0 1 2 3 4 5
Get angry at the actions of other drivers 0 1 2 3 4 5
Drive more than 10km/hr above the speed limit during day-time 0 1 2 3 4 5
Drive after having a few drinks 0 1 2 3 4 5
Enjoy driving 0 1 2 3 4 5
Prefer not to wear a seatbelt 0 1 2 3 4 5
Drive more than 10km/hr above the speed limit at night-time 0 1 2 3 4 5
Feel tired when driving during the day 0 1 2 3 4 5
Enjoy driving faster than other traffic 0 1 2 3 4 5 5. Road Crashes The next few questions are about any road crashes you have been involved in as a Driver in the past 3 years. By crash, we mean any collision that involved either injury to another person or yourself, damage to property, damage to another vehicle, or damage to the vehicle you were driving. Please mention only those crashes that you were involved in as a Driver. How many crashes have you been involved in as a Driver during the past 3 years? (Write number on line)
__________
If you have not been in a crash as a Driver during the past 3 years, please skip to Question 6. Of the crashes that you have been involved in as a Driver during the past 3 years, how many resulted in: Damage only ______________ Slight injury (to any person)
______________
Serious injury (to any person)
______________
Fatality (of any person)
______________
416
Did any of the road crashes occur at either an intersection or a railway level crossing? No ………………………….………… 0 Yes ………………………….………… 1
Accident at an Intersection …………………… Accident at a Railway Level Crossing ……….
1 2
Thinking about your most recent crash (during the past 3 years), what type of collision was it?
(Circle one number for each statement)
Yes No
Head on collision with another vehicle
1 2
Rear-end collision with another vehicle
1 2
Angular (i.e. side-on) collision with another vehicle
1 2
Collision with a pedestrian
1 2
Collision with another object (e.g. parked car, animal, tree)
1 2
Overturned vehicle
1 2
Other
1 2
Was the crash your fault?
Not at all Partially Fully (Circle one number only) 0 1 2
Were the following statements true or false at the time of the crash?
(Circle one number for each statement)
True False
I had been drinking alcohol before driving
1 2
I felt tired
1 2
I was driving too fast
1 2
I was talking to a passenger in the vehicle
1 2
I was listening to music or the radio
1 2
I was talking on a mobile phone
1 2
I was trying to pick up something from the seat or floor
1 2
I was adjusting the radio/cassette/CD or fan/air conditioning
1 2
I was checking the instruments (e.g. fuel gauge or speedometer)
1 2
417
(Circle one number for each statement)
True False
I was daydreaming
1 2
I was distracted by something inside the vehicle
1 2
I was distracted by something outside the vehicle
1 2
6. General Driving Behaviour No one is perfect. Even the best drivers make mistakes, do foolish things, or bend the rules at some time or another. Some of these behaviours are trivial, but some are potentially dangerous. For each item below you are asked to indicate HOW OFTEN, you drive in the following way.. On the scale of 0 ‘Never’ to 5 ‘Nearly all the time’, please circle the number which best sums up your answer.
(Circle one number for each statement) How often do you…..
Nev
er
Har
dly
ever
Occ
asio
nally
Qui
te o
ften
Freq
uent
ly
Nea
rly a
ll th
e tim
e
Attempt to overtake someone that you hadn’t noticed to be signalling a right turn
0 1 2 3 4 5
Get into the wrong lane when approaching a roundabout or a junction
0 1 2 3 4 5
Miss ‘Stop’ or ‘Give Way’ signs and narrowly avoid colliding with traffic having right of way
0 1 2 3 4 5
Misread the signs and exit from the roundabout on the wrong road
0 1 2 3 4 5
Fail to notice that pedestrians are crossing when turning into a side street from a main road
0 1 2 3 4 5
Drive especially close to the car in front as a signal to its driver to go faster or get out of the way
0 1 2 3 4 5
Forget where you left your car in the car park
0 1 2 3 4 5
Queuing to turn left onto a main road, you pay such close attention to the mainstream of traffic that you nearly hit the car in front
0 1 2 3 4 5
Hit something when reversing that you had not previously seen 0 1 2 3 4 5
Cross a junction knowing that the traffic lights have already turned against you
0 1 2 3 4 5
On turning left nearly hit a cyclist who has come up on your inside 0 1 2 3 4 5
Disregard the speed limits late at night or very early in the morning 0 1 2 3 4 5
Attempt to drive away from the traffic lights in third gear 0 1 2 3 4 5
Fail to check your rear-view mirror before pulling out, changing lanes, etc.
0 1 2 3 4 5
418
(Circle one number for each statement) How often do you…..
Nev
er
Har
dly
ever
Occ
asio
nally
Qui
te o
ften
Freq
uent
ly
Nea
rly a
ll th
e tim
e
Have an aversion to a particular class of road user, and indicate your hostility by whatever means you can
0 1 2 3 4 5
Become impatient with a slow driver in the outer lane and overtake on the inside
0 1 2 3 4 5
Underestimate the speed of an oncoming vehicle when overtaking 0 1 2 3 4 5
Switch on one thing, such as the headlights, when you meant to switch on something else, such as the wipers
0 1 2 3 4 5
Brake too quickly on a slippery road, or steer the wrong way in a skid 0 1 2 3 4 5
Intending to drive to destination A, you ‘wake up’ to find yourself on the road to destination B, perhaps because the latter is your more usual destination
0 1 2 3 4 5
Drive even though you realise you may be over the legal blood-alcohol limit
0 1 2 3 4 5
Get involved in unofficial ‘races’ with other drivers 0 1 2 3 4 5
Realise that you have no clear recollection of the road along which you have just been travelling
0 1 2 3 4 5
Angered by another driver’s behaviour, you give chase with the intention of giving him/her a piece of your mind
0 1 2 3 4 5
7. Driving at Railway Level Crossings Please circle how frequently you drive through the following types of railway level crossings when you drive.
(Circle one number for each statement) N
ever
Onc
e a
year
Twic
e a
year
Mon
thly
Wee
kly
Dai
ly
Boom gates with flashing lights
0 1 2 3 4 5
Flashing lights only (no Boom Gate)
0 1 2 3 4 5
Only STOP or GIVE WAY sign
0 1 2 3 4 5
419
How often have you ever engaged in the following behaviours at a Railway Level Crossing? On the scale of 0 ‘Not at all’ to 5 ‘Very often’, please circle the number which best sums up your answer. If you have never driven over any type of railway level crossing (i.e. with or without boom gates), please circle ‘Not Applicable’.
(Circle one number for each statement) I have: N
ot a
t all
Very
O
ften
Not
App
licab
le
Driven through a rail crossing when lights are flashing, but boom gates are yet to descend
0 1 2 3 4 5 n/a
Driven through a rail crossing when boom gates are descending
0 1 2 3 4 5 n/a
Driven through a rail crossing when booms gates are beginning to rise, but lights are still flashing
0 1 2 3 4 5 n/a
Rolling through a rail crossing without stopping, if no train is visible
0 1 2 3 4 5 n/a
Driven through a rail crossing after the first train has passed, without looking for a second
0 1 2 3 4 5 n/a
Failed to look for a train before crossing a rail crossing
0 1 2 3 4 5 n/a
Driven through a rail crossing when the train is visible, but still some distance away
0 1 2 3 4 5 n/a
Driven through a rail crossing when the train is close
0 1 2 3 4 5 n/a
Queued over a rail crossing
0 1 2 3 4 5 n/a
Driven around boom gates to cross a rail crossing
0 1 2 3 4 5 n/a
Tried to beat the train across the rail crossing
0 1 2 3 4 5 n/a
Sped on approach to a rail crossing
0 1 2 3 4 5 n/a
Driven through a rail crossing when the lights are flashing, but no train is visible
0 1 2 3 4 5 n/a
Driven through a rail crossing when the lights are flashing, and the train is visible
0 1 2 3 4 5 n/a
Scanned on approach to a rail crossing, but failed to stop at it
0 1 2 3 4 5 n/a
Stopped on the yellow hatching road markings
0 1 2 3 4 5 n/a
Followed another vehicle across the rail crossing without looking for myself
0 1 2 3 4 5 n/a
Driven through a rail crossing when visibility is impaired
0 1 2 3 4 5 n/a
Driven through a rail crossing without realising it until after the vehicle is over the tracks
0 1 2 3 4 5 n/a
Seen police enforcing the road rules at a rail crossing
0 1 2 3 4 5 n/a
420
We would like to know the likelihood that you will engage in the following behaviours in the next 6 months of driving at Railway Level Crossings. Even if you rarely drive through railway level crossings, please indicate how you would drive if you were to drive through a crossing. On the scale of 0 ‘Not at all likely’ to 5 ‘Very likely’, please circle the number which best describes your answer. (Circle one number for each statement) In the next 6 months of driving, it is likely I will: N
ot a
t all
likel
y
Very
lik
ely
Drive through a rail crossing when lights are flashing, but boom gates are yet to descend
0 1 2 3 4 5
Drive through a rail crossing when boom gates are descending
0 1 2 3 4 5
Drive through a rail crossing when booms gates are beginning to rise, but lights are still flashing
0 1 2 3 4 5
Roll through a rail crossing without stopping, if no train is visible
0 1 2 3 4 5
Drive through a rail crossing after the first train has passed, without looking for a second
0 1 2 3 4 5
Fail to look for a train before crossing a rail crossing
0 1 2 3 4 5
Drive through a rail crossing when the train is visible, but still some distance away
0 1 2 3 4 5
Drive through a rail crossing when the train is close
0 1 2 3 4 5
Queue over a rail crossing
0 1 2 3 4 5
Drive around boom gates to cross
0 1 2 3 4 5
Try to beat the train across the rail crossing
0 1 2 3 4 5
Speed on approach to a rail crossing
0 1 2 3 4 5
Drive through a rail crossing when the lights are flashing, but no train is visible
0 1 2 3 4 5
Drive through a rail crossing when the lights are flashing, and the train is visible
0 1 2 3 4 5
Scan on approach to a rail crossing, but fail to stop at it
0 1 2 3 4 5
Stop on the yellow hatching road markings
0 1 2 3 4 5
Follow another vehicle across the rail crossing without looking for myself
0 1 2 3 4 5
Drive through a rail crossing when visibility is impaired
0 1 2 3 4 5
Drive through a rail crossing without realising until after the vehicle is over the tracks
0 1 2 3 4 5
See police enforcing the road rules at a rail crossing
0 1 2 3 4 5
421
Rate each of the following statements according to the degree they reflect your personal views. On the scale of ‘1’ ‘Strongly Disagree’ to ‘6’ ‘Strongly Agree’, please circle the number which best describes your answer. (Circle one number for each statement)
Stro
ngly
D
isag
ree
Stro
ngly
A
gree
Your family generally obeys the rules at rail crossings 1 2 3 4 5 6
Your friends generally obey the rules at rail crossings 1 2 3 4 5 6
Other motorists generally obey the rules at rail crossings 1 2 3 4 5 6
Your family generally think it important to obey the rules at rail crossings 1 2 3 4 5 6
Your friends generally think it important to obey the rules at rail crossings 1 2 3 4 5 6
Other motorists generally think it important to obey the rules at rail crossings
1 2 3 4 5 6
It is generally safe to disobey the rules at rail crossings 1 2 3 4 5 6
It is generally possible to judge a train’s speed 1 2 3 4 5 6
It is generally safe to cross if you can’t see a train, even if the lights are flashing
1 2 3 4 5 6
It is generally safe to roll slowly through a crossing instead of stopping 1 2 3 4 5 6
Trains generally run to a regular timetable 1 2 3 4 5 6
Penalties need to be tougher for violating road rules at rail crossings 1 2 3 4 5 6
The main deterrent for breaking the rules at rail crossings is fear of getting caught
1 2 3 4 5 6
Generally it is more important to use common sense at rail crossings than strictly follow the road rules
1 2 3 4 5 6
We would like to ask you some questions about your attitude towards railway level crossings. Please circle one number in each table. As a motorist, I believe that the design of rail crossings are:
Bad Good -3 -2 -1 0 1 2 3
Unsafe Safe
-3 -2 -1 0 1 2 3
As a motorist, I believe that the design of rail crossings are:
Confusing Easily Understood
-3 -2 -1 0 1 2 3
422
As a motorist, I believe that the design of rail crossings are:
Difficult for obeying
road rules
Easy for obeying
road rules -3 -2 -1 0 1 2 3
As a motorist, I believe that road rules at rail crossings are:
Bad
Good
-3 -2 -1 0 1 2 3
Not Strict Enough
Too Strict
-3 -2 -1 0 1 2 3
As a motorist, I believe that road rules at rail crossings are:
Confusing
Easily Understood
-3 -2 -1 0 1 2 3
Not
Practical
Practical -3 -2 -1 0 1 2 3
Obeying the road rules at rail crossings is:
Not up to me
Up to me
-3 -2 -1 0 1 2 3
Out of my
control Under my
control -3 -2 -1 0 1 2 3
Obeying the road rules at rail crossings is:
Dependent on other
motorists
Dependent on me only
-3 -2 -1 0 1 2 3
423
Obeying the road rules at rail crossings is:
Dependent on time
constraints
Not dependent
on time constraints
-3 -2 -1 0 1 2 3
Other motorists influence my driving behaviour at rail crossings, making it:
Harder to obey road
rules
Easier to obey road
rules -3 -2 -1 0 1 2 3
More
confusing Less
confusing -3 -2 -1 0 1 2 3
More
stressful More
relaxing -3 -2 -1 0 1 2 3
When you are driving at a Railway Level Crossing, how often do each of the following occur? On the scale of 0 ‘Never’ to ‘5’ ‘Always’, please circle the number which best describes your answer.
(Circle one number for each statement) N
ever
Alw
ays
Blinding sun makes it difficult to see if the red flashing lights are activated
0 1 2 3 4 5
The design of the road makes it difficult to see if a train is approaching or at the rail crossing
0 1 2 3 4 5
Warning systems on the road approaching the crossing are not adequate to inform drivers there is a rail crossing ahead
0 1 2 3 4 5
Road surfaces are often poor and it is difficult to stop
0 1 2 3 4 5
Boom gates and/or flashing lights are often faulty
0 1 2 3 4 5
Other drivers do stupid things that put you in a dangerous situation
0 1 2 3 4 5
Intersections ahead of a rail crossing often cause your car to overhang the tracks
0 1 2 3 4 5
Difficult to hear an approaching train when the windows are up
0 1 2 3 4 5
My car stalled on the tracks at a rail crossing
0 1 2 3 4 5
424
The following section tests your knowledge about railway level crossings. Circle the number that best describes your answer. When the twin red lights start flashing this means?
A train is coming but it is still safe to cross………….. You must stop and not enter the crossing ………….. Don’t know ………………………………………………
1 2 3
Boom gates usually stay down a lot longer than traffic lights take to change at an intersection
True ………………………………………………………False ……………........…………………………………. Don’t know ………………………………………………
1 2 3
Yellow hatching road markings means keep clear
True ………………………………………………………False …………………………………………………….. Don’t know ………………………………………………
1 2 3
There are fines for not stopping at railway crossings
True ………………………………………………………False …………………………………………………….. Don’t know ………………………………………………
1 2 3
How likely is it, that as a Driver you will be involved in a crash with a train at a railway level crossing?
Not at all Likely
Very Likely
0 1 2 3 4 5
We would like to be able to match this questionnaire with others at a later date. However, we wish to preserve your anonymity. One way to do this is to collect some information from you that will not identify you, but will allow us to match the questionnaires. Please assist us by providing:
First 3 letters of your mother’s maiden name (i.e. Surname before marriage): What is your Date of Birth? _______________________ (e.g. 3rd July 1986)
Thank you for completing this questionnaire. Please return this questionnaire in
the reply-paid envelope supplied.
425
ID Number: __ __ __ - __ Staff Use Only
Centre for Accident Research and Road Safety – Queensland
Young Drivers Post-Test Questionnaire
November 2006
Associate Professor Jeremy Davey Deputy Director Centre for Accident Research & Road Safety – Queensland Queensland University of Technology Telephone: 07 3864 4574 Email: [email protected]
Angela Wallace PhD Scholar Centre for Accident Research & Road Safety – Queensland Queensland University of Technology Telephone: 0402 240 234 Email: [email protected]
426
General Driving Behaviour No one is perfect. Even the best drivers make mistakes, do foolish things, or bend the rules at some time or another. Some of these behaviours are trivial, but some are potentially dangerous. For each item below you are asked to indicate HOW OFTEN, if at all, this kind of thing has happened to you. On the scale of 0 ‘Never’ to 5 ‘Nearly all the time’, please circle the number which best sums up your answer.
(Circle one number for each statement) How often do you…..
Nev
er
Har
dly
ever
Occ
asio
nally
Qui
te o
ften
Freq
uent
ly
Nea
rly a
ll th
e tim
e
Attempt to overtake someone that you hadn’t noticed to be signalling a right turn
0 1 2 3 4 5
Get into the wrong lane when approaching a roundabout or a junction
0 1 2 3 4 5
Miss ‘Stop’ or ‘Give Way’ signs and narrowly avoid colliding with traffic having right of way
0 1 2 3 4 5
Misread the signs and exit from the roundabout on the wrong road
0 1 2 3 4 5
Fail to notice that pedestrians are crossing when turning into a side street from a main road
0 1 2 3 4 5
Drive especially close to the car in front as a signal to its driver to go faster or get out of the way
0 1 2 3 4 5
Forget where you left your car in the car park
0 1 2 3 4 5
Queuing to turn left onto a main road, you pay such close attention to the mainstream of traffic that you nearly hit the car in front
0 1 2 3 4 5
Hit something when reversing that you had not previously seen 0 1 2 3 4 5
Cross a junction knowing that the traffic lights have already turned against you
0 1 2 3 4 5
On turning left nearly hit a cyclist who has come up on your inside 0 1 2 3 4 5
Disregard the speed limits late at night or very early in the morning 0 1 2 3 4 5
Attempt to drive away from the traffic lights in third gear 0 1 2 3 4 5
Fail to check your rear-view mirror before pulling out, changing lanes, etc.
0 1 2 3 4 5
Have an aversion to a particular class of road user, and indicate your hostility by whatever means you can
0 1 2 3 4 5
Become impatient with a slow driver in the outer lane and overtake on the inside
0 1 2 3 4 5
Underestimate the speed of an oncoming vehicle when overtaking 0 1 2 3 4 5
Switch on one thing, such as the headlights, when you meant to switch on something else, such as the wipers
0 1 2 3 4 5
427
(Circle one number for each statement) How often do you…..
Nev
er
Har
dly
ever
Occ
asio
nally
Qui
te o
ften
Freq
uent
ly
Nea
rly a
ll th
e tim
e
Brake too quickly on a slippery road, or steer the wrong way in a skid 0 1 2 3 4 5
Intending to drive to destination A, you ‘wake up’ to find yourself on the road to destination B, perhaps because the latter is your more usual destination
0 1 2 3 4 5
Drive even though you realise you may be over the legal blood-alcohol limit
0 1 2 3 4 5
Get involved in unofficial ‘races’ with other drivers 0 1 2 3 4 5
Realise that you have no clear recollection of the road along which you have just been travelling
0 1 2 3 4 5
Angered by another driver’s behaviour, you give chase with the intention of giving him/her a piece of your mind
0 1 2 3 4 5
Driving at Railway Level Crossings How often have you engaged in the following behaviours whilst driving at a railway level crossing in the past 2-3 weeks? On the scale of 0 ‘Not at all’ to 5 ‘Very often’, please circle the number which best sums up your answer. If you have not driven over any type of railway crossing during the past 2-3 weeks, please circle ‘Not Applicable’.
(Circle one number for each statement) I have: N
ot a
t all
Very
O
ften
Not
App
licab
le
Driven through a rail crossing when lights are flashing, but boom gates are yet to descend
0 1 2 3 4 5 n/a
Driven through a rail crossing when boom gates are descending
0 1 2 3 4 5 n/a
Driven through a rail crossing when booms gates are beginning to rise, but lights are still flashing
0 1 2 3 4 5 n/a
Rolling through a rail crossing without stopping, if no train is visible
0 1 2 3 4 5 n/a
Driven through a rail crossing after the first train has passed, without looking for a second
0 1 2 3 4 5 n/a
Failed to look for a train before crossing a rail crossing
0 1 2 3 4 5 n/a
Driven through a rail crossing when the train is visible, but still some distance away
0 1 2 3 4 5 n/a
Driven through a rail crossing when the train is close
0 1 2 3 4 5 n/a
428
(Circle one number for each statement) I have: N
ot a
t all
Very
O
ften
Not
App
licab
le
Queued over a rail crossing
0 1 2 3 4 5 n/a
Driven around boom gates to cross a rail crossing
0 1 2 3 4 5 n/a
Tried to beat the train across the rail crossing
0 1 2 3 4 5 n/a
Sped on approach to a rail crossing
0 1 2 3 4 5 n/a
Driven through a rail crossing when the lights are flashing, but no train is visible
0 1 2 3 4 5 n/a
Driven through a rail crossing when the lights are flashing, and the train is visible
0 1 2 3 4 5 n/a
Scanned on approach to a rail crossing, but failed to stop at it
0 1 2 3 4 5 n/a
Stopped on the yellow hatching road markings
0 1 2 3 4 5 n/a
Followed another vehicle across the rail crossing without looking for myself
0 1 2 3 4 5 n/a
Driven through a rail crossing when visibility is impaired
0 1 2 3 4 5 n/a
Driven through a rail crossing without realising it until after the vehicle is over the tracks
0 1 2 3 4 5 n/a
Seen police enforcing the road rules at a rail crossing
0 1 2 3 4 5 n/a
We would like to know the likelihood that you will engage in the following behaviours in the next 6 months of driving at railway level crossings. Even if you rarely drive through railway level crossings, please indicate how you would drive if you were to drive through a railway crossing. On the scale of 0 ‘Not at all likely’ to 5 ‘Very likely’, please circle the number which best describes your answer. (Circle one number for each statement) In the next 6 months of driving, it is likely I will: N
ot a
t all
likel
y
Ve
ry
likel
y
Drive through a rail crossing when lights are flashing, but boom gates are yet to descend
0 1 2 3 4 5
Drive through a rail crossing when boom gates are descending
0 1 2 3 4 5
Drive through a rail crossing when booms gates are beginning to rise, but lights are still flashing
0 1 2 3 4 5
Roll through a rail crossing without stopping, if no train is visible
0 1 2 3 4 5
Drive through a rail crossing after the first train has passed, without looking for a second
0 1 2 3 4 5
Fail to look for a train before crossing a rail crossing
0 1 2 3 4 5
429
(Circle one number for each statement) In the next 6 months of driving, it is likely I will: N
ot a
t all
Like
ly
Very
Li
kely
Drive through a rail crossing when the train is visible, but still some distance away
0 1 2 3 4 5
Drive through a rail crossing when the train is close
0 1 2 3 4 5
Queue over a rail crossing
0 1 2 3 4 5
Drive around boom gates to cross
0 1 2 3 4 5
Try to beat the train across the rail crossing
0 1 2 3 4 5
Speed on approach to a rail crossing
0 1 2 3 4 5
Drive through a rail crossing when the lights are flashing, but no train is visible
0 1 2 3 4 5
Drive through a rail crossing when the lights are flashing, and the train is visible
0 1 2 3 4 5
Scan on approach to a rail crossing, but fail to stop at it
0 1 2 3 4 5
Stop on the yellow hatching road markings
0 1 2 3 4 5
Follow another vehicle across the rail crossing without looking for myself
0 1 2 3 4 5
Drive through a rail crossing when visibility is impaired
0 1 2 3 4 5
Drive through a rail crossing without realising until after the vehicle is over the tracks
0 1 2 3 4 5
See police enforcing the road rules at a rail crossing
0 1 2 3 4 5
Rate each of the following statements according to the degree they reflect your personal views. On the scale of ‘1’ ‘Strongly Disagree’ to ‘6’ ‘Strongly Agree’, please circle the number which best describes your answer. (Circle one number for each statement)
Stro
ngly
D
isag
ree
St
rong
ly
Agr
ee
Your family generally obeys the rules at rail crossings 1 2 3 4 5 6
Your friends generally obey the rules at rail crossings 1 2 3 4 5 6
Other motorists generally obey the rules at rail crossings 1 2 3 4 5 6
Your family generally think it important to obey the rules at rail crossings 1 2 3 4 5 6
Your friends generally think it important to obey the rules at rail crossings 1 2 3 4 5 6
Other motorists generally think it important to obey the rules at rail crossings
1 2 3 4 5 6
It is generally safe to disobey the rules at rail crossings 1 2 3 4 5 6
It is generally possible to judge a train’s speed 1 2 3 4 5 6
430
(Circle one number for each statement)
Stro
ngly
D
isag
ree
Stro
ngly
A
gree
It is generally safe to cross if you can’t see a train, even if the lights are flashing
1 2 3 4 5 6
It is generally safe to roll slowly through a crossing instead of stopping 1 2 3 4 5 6
Trains generally run to a regular timetable 1 2 3 4 5 6
Penalties need to be tougher for violating road rules at rail crossings 1 2 3 4 5 6
The main deterrent for breaking the rules at rail crossings is fear of getting caught
1 2 3 4 5 6
Generally it is more important to use common sense at rail crossings than strictly follow the road rules
1 2 3 4 5 6
We would like to ask you some questions about your attitude towards railway level crossings. Please circle one number in each table. As a motorist, I believe that the design of rail crossings are:
Bad Good -3 -2 -1 0 1 2 3
Unsafe Safe -3 -2 -1 0 1 2 3
Confusing Easily Understood
-3 -2 -1 0 1 2 3
Difficult for obeying
road rules
Easy for obeying
road rules -3 -2 -1 0 1 2 3
As a motorist, I believe that road rules at rail crossings are:
Bad
Good
-3 -2 -1 0 1 2 3
Not Strict Enough
Too Strict
-3 -2 -1 0 1 2 3
431
As a motorist, I believe that road rules at rail crossings are:
Confusing
Easily Understood
-3 -2 -1 0 1 2 3
Not
Practical
Practical -3 -2 -1 0 1 2 3
Obeying the road rules at rail crossings is:
Not up to me
Up to me
-3 -2 -1 0 1 2 3
Out of my
control Under my
control -3 -2 -1 0 1 2 3
Obeying the road rules at rail crossings is:
Dependent on other
motorists
Dependent on me only
-3 -2 -1 0 1 2 3
Obeying the road rules at rail crossings is:
Dependent
on time constraints
Not dependent
on time constraints
-3 -2 -1 0 1 2 3
Other motorists influence my driving behaviour at rail crossings, making it:
Harder to obey road
rules
Easier to obey road
rules -3 -2 -1 0 1 2 3
More
confusing Less
confusing -3 -2 -1 0 1 2 3
More
stressful More
relaxing -3 -2 -1 0 1 2 3
432
When you are driving at railway level crossings, how often do each of the following occur? On the scale of 0 ‘Never’ to ‘5’ ‘Always’, please circle the number which best describes your answer.
(Circle one number for each statement) N
ever
Alw
ays
Blinding sun makes it difficult to see if the red flashing lights are activated
0 1 2 3 4 5
The design of the road makes it difficult to see if a train is approaching or at the rail crossing
0 1 2 3 4 5
Warning systems on the road approaching the crossing are not adequate to inform drivers there is a rail crossing ahead
0 1 2 3 4 5
Road surfaces are often poor and it is difficult to stop
0 1 2 3 4 5
Boom gates and/or flashing lights are often faulty
0 1 2 3 4 5
Other drivers do stupid things that put you in a dangerous situation
0 1 2 3 4 5
Intersections ahead of a rail crossing often cause your car to overhang the tracks
0 1 2 3 4 5
Difficult to hear an approaching train when the windows are up 0 1 2 3 4 5 My car stalled on the tracks at a rail crossing
0 1 2 3 4 5
The following section tests your knowledge about railway level crossings. Circle the number that best describes your answer. When the twin red lights start flashing this means?
A train is coming but it is still safe to cross………….. You must stop and not enter the crossing ………….. Don’t know ………………………………………………
1 2 3
Boom gates usually stay down a lot longer than traffic lights take to change at an intersection
True ………………………………………………………False ……………........…………………………………. Don’t know ………………………………………………
1 2 3
Yellow hatching road markings means keep clear
True ………………………………………………………False …………………………………………………….. Don’t know ………………………………………………
1 2 3
There are fines for not stopping at railway crossings
True ………………………………………………………False …………………………………………………….. Don’t know ………………………………………………
1 2 3
433
How likely is it, that as a Driver you will be involved in a crash with a train at a railway level crossing?
Not at all Likely
Very Likely
0 1 2 3 4 5
Your Date of Birth: __________________________________ (e.g. 3rd July 1986)
Thank you for completing this questionnaire. Please return this questionnaire in the reply-paid envelope supplied.
434
ID Number: __ __ __ - __ Staff Use Only
Centre for Accident Research and Road Safety – Queensland
Drivers 60 + years Pre-Test Questionnaire
August 2006
Associate Professor Jeremy Davey Deputy Director Centre for Accident Research & Road Safety – Queensland Queensland University of Technology Telephone: 07 3864 4574 Email: [email protected]
Angela Wallace PhD Scholar Centre for Accident Research & Road Safety – Queensland Queensland University of Technology Telephone: 0402 240 234 Email: [email protected]
435
1. Personal Information We would like to start by asking you some questions about yourself. Please circle the number that best describes you.
What is your gender? Male………………………………………………….……
Female…………………………………………………… 1 2
What is your age? _________ Years
Are you currently involved in any paid work?
Yes, I am currently working …………………………… No, I am not currently working ………………………..
1 2
If you are in paid work, what is your occupation?
_______________________________________
What is the postcode where you live?
Do you have any conditions on your Drivers Licence?
No conditions….………………………………………… ‘S’ condition only (corrective lenses required)…….… ‘M’ condition only (medical certificate required) …….. Both ‘S’ & ‘M’ conditions ……………………………….
1 2 3 4
2. Driving Patterns We would like to ask you some questions about how often you drive.
On average, how many days of the week do you drive? ______________________ days per week
On average, how many hours per day do you drive? ______________________ hours per day 3. Health We would now like to ask you some questions about your health. Have you ever had, or been told by a Doctor that you have any of the following medical conditions?
(Circle one number for each statement)
Yes No
High Blood Pressure 1 2
Heart Disease 1 2
Chest Pain / Angina 1 2
Any condition requiring Heart Surgery 1 2
Palpitations / Irregular Heart Beat 1 2
Head Injury / Spinal Injury 1 2
Seizures, Fits, Convulsions or Epilepsy 1 2
Abnormal Shortness of Breath 1 2
Blackouts or Fainting 1 2
Stroke 1 2
436
(Circle one number for each statement)
Yes No
Dizziness, Vertigo or Problems with Balance 1 2
Double Vision or Difficulty Seeing 1 2
Colour Blindness 1 2
Kidney Disease 1 2
Diabetes 1 2
Neck, Back or Limb Disorders 1 2
Hearing Loss or Deafness, or had an Ear Operation or a Hearing Aid 1 2
4. Driving Self-Assessment
The next few questions relate to your driving. We would like for you to answer these questions as truthfully as possible. (Circle one number for each statement) N
ot a
t al
l
Very
O
ften
Are your reactions to unexpected situations slower than they used to be?
0 1 2 3 4 5
Do you have trouble judging the distance of other vehicles, or changing focus from your instrument panel to the road?
0 1 2 3 4 5
Are you having more trouble adjusting to glare and/or night driving than you did previously?
0 1 2 3 4 5
Do you ever get surprised by pedestrians or other vehicles coming from your left or right while you are focusing straight ahead?
0 1 2 3 4 5
Do some traffic situations or other drivers upset you?
0 1 2 3 4 5
Do you have trouble driving through, or turning at busy intersections or roundabouts?
0 1 2 3 4 5
Do you feel uncomfortable driving in unfamiliar territory?
0 1 2 3 4 5
Do you find that you are easily distracted or that your thoughts wander while you are driving?
0 1 2 3 4 5
Do you have regular health and vision checks?
0 1 2 3 4 5
5. Road Crashes The next few questions are about any road crashes you have been involved in as a Driver in the past 3 years. By crash, we mean any collision that involved either injury to another person or yourself, damage to property, damage to another vehicle, or damage to the vehicle you were driving. Please mention only those crashes that you were involved in as a Driver. How many crashes have you been involved in as a Driver during the past 3 years? (Write number on line)
__________
437
If you have not been in a crash as a Driver during the past 3 years, please skip to Question 6. Of the crashes that you have been involved in as a Driver during the past 3 years, how many resulted in: Damage only ______________ Slight injury (to any person)
______________
Serious injury (to any person)
______________
Fatality (of any person)
______________
Did any of the road crashes occur at either an intersection or a railway level crossing? No ………………………….………… 0 Yes ………………………….………… 1
Accident at an Intersection …………………… Accident at a Railway Level Crossing ……….
1 2
Thinking about your most recent crash (during the past 3 years), what type of collision was it?
(Circle one number for each statement)
Yes No
Head on collision with another vehicle
1 2
Rear-end collision with another vehicle
1 2
Angular (i.e. side-on) collision with another vehicle
1 2
Collision with a pedestrian
1 2
Collision with another object (e.g. parked car, animal, tree)
1 2
Overturned vehicle
1 2
Other
1 2
Was the crash your fault?
Not at all Partially Fully (Circle one number only) 0 1 2
Were the following statements true or false at the time of the crash?
(Circle one number for each statement)
True False
I had been drinking alcohol before driving
1 2
I felt tired
1 2
I was driving too fast
1 2
I was talking to a passenger in the vehicle 1 2
438
(Circle one number for each statement)
True False
I was listening to music or the radio
1 2
I was talking on a mobile phone
1 2
I was trying to pick up something from the seat or floor
1 2
I was adjusting the radio/cassette/CD or fan/air conditioning
1 2
I was checking the instruments (e.g. fuel gauge or speedometer)
1 2
I was daydreaming
1 2
I was distracted by something inside the vehicle
1 2
I was distracted by something outside the vehicle
1 2
6. General Driving Behaviour No one is perfect. Even the best drivers make mistakes, do foolish things, or bend the rules at some time or another. Some of these behaviours are trivial, but some are potentially dangerous. For each item below you are asked to indicate HOW OFTEN, if at all, this kind of thing has happened to you. On the scale of 0 ‘Never’ to 5 ‘Nearly all the time’, please circle the number which best sums up your answer.
(Circle one number for each statement) How often do you…..
Nev
er
Har
dly
ever
Occ
asio
nally
Qui
te o
ften
Freq
uent
ly
Nea
rly a
ll th
e tim
e
Attempt to overtake someone that you hadn’t noticed to be signalling a right turn
0 1 2 3 4 5
Get into the wrong lane when approaching a roundabout or a junction
0 1 2 3 4 5
Miss ‘Stop’ or ‘Give Way’ signs and narrowly avoid colliding with traffic having right of way
0 1 2 3 4 5
Misread the signs and exit from the roundabout on the wrong road
0 1 2 3 4 5
Fail to notice that pedestrians are crossing when turning into a side street from a main road
0 1 2 3 4 5
Drive especially close to the car in front as a signal to its driver to go faster or get out of the way
0 1 2 3 4 5
Forget where you left your car in the car park
0 1 2 3 4 5
Queuing to turn left onto a main road, you pay such close attention to the mainstream of traffic that you nearly hit the car in front
0 1 2 3 4 5
Hit something when reversing that you had not previously seen 0 1 2 3 4 5
439
(Circle one number for each statement) How often do you…..
Nev
er
Har
dly
ever
Occ
asio
nally
Qui
te o
ften
Freq
uent
ly
Nea
rly a
ll th
e tim
e
Cross a junction knowing that the traffic lights have already turned against you
0 1 2 3 4 5
On turning left nearly hit a cyclist who has come up on your inside 0 1 2 3 4 5
Disregard the speed limits late at night or very early in the morning 0 1 2 3 4 5
Attempt to drive away from the traffic lights in third gear 0 1 2 3 4 5
Fail to check your rear-view mirror before pulling out, changing lanes, etc.
0 1 2 3 4 5
Have an aversion to a particular class of road user, and indicate your hostility by whatever means you can
0 1 2 3 4 5
Become impatient with a slow driver in the outer lane and overtake on the inside
0 1 2 3 4 5
Underestimate the speed of an oncoming vehicle when overtaking 0 1 2 3 4 5
Switch on one thing, such as the headlights, when you meant to switch on something else, such as the wipers
0 1 2 3 4 5
Brake too quickly on a slippery road, or steer the wrong way in a skid 0 1 2 3 4 5
Intending to drive to destination A, you ‘wake up’ to find yourself on the road to destination B, perhaps because the latter is your more usual destination
0 1 2 3 4 5
Drive even though you realise you may be over the legal blood-alcohol limit
0 1 2 3 4 5
Get involved in unofficial ‘races’ with other drivers 0 1 2 3 4 5
Realise that you have no clear recollection of the road along which you have just been travelling
0 1 2 3 4 5
Angered by another driver’s behaviour, you give chase with the intention of giving him/her a piece of your mind
0 1 2 3 4 5
440
7. Driving at Railway Level Crossings Please circle how frequently you drive through the following types of railway level crossings when you drive.
(Circle one number for each statement) N
ever
Onc
e a
year
Twic
e a
year
Mon
thly
Wee
kly
Dai
ly
Boom gates with flashing lights
0 1 2 3 4 5
Flashing lights only (no Boom Gate)
0 1 2 3 4 5
Only STOP or GIVE WAY sign
0 1 2 3 4 5
How often have you ever engaged in the following behaviours at a railway level crossing? On the scale of 0 ‘Not at all’ to 5 ‘Very often’, please circle the number which best sums up your answer. If you have never driven over any type of railway level crossing (i.e. with or without boom gates), please circle ‘Not Applicable’.
(Circle one number for each statement) I have: N
ot a
t all
Very
O
ften
Not
App
licab
le
Driven through a rail crossing when lights are flashing, but boom gates are yet to descend
0 1 2 3 4 5 n/a
Driven through a rail crossing when boom gates are descending
0 1 2 3 4 5 n/a
Driven through a rail crossing when booms gates are beginning to rise, but lights are still flashing
0 1 2 3 4 5 n/a
Rolling through a rail crossing without stopping, if no train is visible
0 1 2 3 4 5 n/a
Driven through a rail crossing after the first train has passed, without looking for a second
0 1 2 3 4 5 n/a
Failed to look for a train before crossing a rail crossing
0 1 2 3 4 5 n/a
Driven through a rail crossing when the train is visible, but still some distance away
0 1 2 3 4 5 n/a
Driven through a rail crossing when the train is close
0 1 2 3 4 5 n/a
Queued over a rail crossing
0 1 2 3 4 5 n/a
Driven around boom gates to cross a rail crossing
0 1 2 3 4 5 n/a
Tried to beat the train across the rail crossing
0 1 2 3 4 5 n/a
Sped on approach to a rail crossing
0 1 2 3 4 5 n/a
Driven through a rail crossing when the lights are flashing, but no train is visible
0 1 2 3 4 5 n/a
441
(Circle one number for each statement) I have: N
ot a
t all
Very
O
ften
Not
App
licab
le
Driven through a rail crossing when the lights are flashing, and the train is visible
0 1 2 3 4 5 n/a
Scanned on approach to a rail crossing, but failed to stop at it
0 1 2 3 4 5 n/a
Stopped on the yellow hatching road markings
0 1 2 3 4 5 n/a
Followed another vehicle across the rail crossing without looking for myself
0 1 2 3 4 5 n/a
Driven through a rail crossing when visibility is impaired
0 1 2 3 4 5 n/a
Driven through a rail crossing without realising it until after the vehicle is over the tracks
0 1 2 3 4 5 n/a
Seen police enforcing the road rules at a rail crossing
0 1 2 3 4 5 n/a
We would like to know the likelihood that you will engage in the following behaviours in the next 6 months of driving at railway level crossings. Even if you rarely drive through railway level crossings, please indicate how you would drive if you were to drive through a crossing. On the scale of 0 ‘Not at all likely’ to 5 ‘Very likely’, please circle the number which best describes your answer. (Circle one number for each statement) In the next 6 months of driving, it is likely I will: N
ot a
t all
likel
y
Very
lik
ely
Drive through a rail crossing when lights are flashing, but boom gates are yet to descend
0 1 2 3 4 5
Drive through a rail crossing when boom gates are descending
0 1 2 3 4 5
Drive through a rail crossing when booms gates are beginning to rise, but lights are still flashing
0 1 2 3 4 5
Roll through a rail crossing without stopping, if no train is visible
0 1 2 3 4 5
Drive through a rail crossing after the first train has passed, without looking for a second
0 1 2 3 4 5
Fail to look for a train before crossing a rail crossing
0 1 2 3 4 5
Drive through a rail crossing when the train is visible, but still some distance away
0 1 2 3 4 5
Drive through a rail crossing when the train is close
0 1 2 3 4 5
Queue over a rail crossing
0 1 2 3 4 5
Drive around boom gates to cross
0 1 2 3 4 5
Try to beat the train across the rail crossing
0 1 2 3 4 5
Speed on approach to a rail crossing 0 1 2 3 4 5
442
(Circle one number for each statement) In the next 6 months of driving, it is likely I will: N
ot a
t all
Like
ly
Very
Li
kely
Drive through a rail crossing when the lights are flashing, but no train is visible
0 1 2 3 4 5
Drive through a rail crossing when the lights are flashing, and the train is visible
0 1 2 3 4 5
Scan on approach to a rail crossing, but fail to stop at it
0 1 2 3 4 5
Stop on the yellow hatching road markings
0 1 2 3 4 5
Follow another vehicle across the rail crossing without looking for myself
0 1 2 3 4 5
Drive through a rail crossing when visibility is impaired
0 1 2 3 4 5
Drive through a rail crossing without realising until after the vehicle is over the tracks
0 1 2 3 4 5
See police enforcing the road rules at a rail crossing
0 1 2 3 4 5
Rate each of the following statements according to the degree they reflect your personal views. On the scale of ‘1’ ‘Strongly Disagree’ to ‘6’ ‘Strongly Agree’, please circle the number which best describes your answer.
(Circle one number for each statement)
Stro
ngly
A
gree
Stro
ngly
D
isag
ree
Your family generally obeys the rules at rail crossings 1 2 3 4 5
Your friends generally obey the rules at rail crossings 1 2 3 4 5
Other motorists generally obey the rules at rail crossings 1 2 3 4 5
Your family generally think it important to obey the rules at rail crossings 1 2 3 4 5
Your friends generally think it important to obey the rules at rail crossings 1 2 3 4 5
Other motorists generally think it important to obey the rules at rail crossings
1 2 3 4 5
It is generally safe to disobey the rules at rail crossings 1 2 3 4 5
It is generally possible to judge a train’s speed 1 2 3 4 5
It is generally safe to cross if you can’t see a train, even if the lights are flashing
1 2 3 4 5
It is generally safe to roll slowly through a crossing instead of stopping 1 2 3 4 5
Trains generally run to a regular timetable 1 2 3 4 5
Penalties need to be tougher for violating road rules at rail crossings 1 2 3 4 5
The main deterrent for breaking the rules at rail crossings is fear of getting caught
1 2 3 4 5
Generally it is more important to use common sense at rail crossings than strictly follow the road rules
1 2 3 4 5
443
We would like to ask you some questions about your attitude towards railway level crossings. Please circle one number in each table. As a motorist, I believe that the design of rail crossings are:
Bad Good -3 -2 -1 0 1 2 3
Unsafe Safe
-3 -2 -1 0 1 2 3
As a motorist, I believe that the design of rail crossings are:
Confusing Easily Understood
-3 -2 -1 0 1 2 3
Difficult for
obeying road rules
Easy for obeying
road rules -3 -2 -1 0 1 2 3
As a motorist, I believe that road rules at rail crossings are:
Bad
Good
-3 -2 -1 0 1 2 3
Not Strict Enough
Too Strict
-3 -2 -1 0 1 2 3
As a motorist, I believe that road rules at rail crossings are:
Confusing
Easily Understood
-3 -2 -1 0 1 2 3
Not
Practical
Practical -3 -2 -1 0 1 2 3
Obeying the road rules at rail crossings is:
Not up to me
Up to me
-3 -2 -1 0 1 2 3
Out of my
control Under my
control -3 -2 -1 0 1 2 3
444
Obeying the road rules at rail crossings is:
Dependent on other
motorists
Dependent on me only
-3 -2 -1 0 1 2 3
Obeying the road rules at rail crossings is:
Dependent
on time constraints
Not dependent
on time constraints
-3 -2 -1 0 1 2 3
Other motorists influence my driving behaviour at rail crossings, making it:
Harder to obey road
rules
Easier to obey road
rules -3 -2 -1 0 1 2 3
More
confusing Less
confusing -3 -2 -1 0 1 2 3
More
stressful More
relaxing -3 -2 -1 0 1 2 3
When you are driving a car at railway level crossings, how often do each of the following occur? On the scale of 0 ‘Never’ to ‘5’ ‘Always’, please circle the number which best describes your answer.
(Circle one number for each statement) N
ever
Alw
ays
Blinding sun makes it difficult to see if the red flashing lights are activated
0 1 2 3 4 5
The design of the road makes it difficult to see if a train is approaching or at the rail crossing
0 1 2 3 4 5
Warning systems on the road approaching the crossing are not adequate to inform drivers there is a rail crossing ahead
0 1 2 3 4 5
Road surfaces are often poor and it is difficult to stop
0 1 2 3 4 5
Boom gates and/or flashing lights are often faulty
0 1 2 3 4 5
Other drivers do stupid things that put you in a dangerous situation
0 1 2 3 4 5
Intersections ahead of a rail crossing often cause your car to overhang the tracks
0 1 2 3 4 5
Difficult to hear an approaching train when the windows are up 0 1 2 3 4 5
445
(Circle one number for each statement) N
ever
Alw
ays
My car stalled on the tracks at a rail crossing
0 1 2 3 4 5
The following section tests your knowledge about railway level crossings. Circle the number that best describes your answer. When the twin red lights start flashing this means?
A train is coming but it is still safe to cross………….. You must stop and not enter the crossing ………….. Don’t know ………………………………………………
1 2 3
Boom gates usually stay down a lot longer than traffic lights take to change at an intersection
True ………………………………………………………False ……………........…………………………………. Don’t know ………………………………………………
1 2 3
Yellow hatching road markings means keep clear
True ………………………………………………………False …………………………………………………….. Don’t know ………………………………………………
1 2 3
There are fines for not stopping at railway crossings
True ………………………………………………………False …………………………………………………….. Don’t know ………………………………………………
1 2 3
How likely is it, that as a Driver you will be involved in a crash with a train at a railway level crossing?
Not at all Likely
Very Likely
0 1 2 3 4 5
We would like to be able to match this questionnaire with others at a later date. However, we wish to preserve your anonymity. One way to do this is to collect some information from you that will not identify you, but will allow us to match the questionnaires. Please assist us by providing:
First 3 letters of your mother’s maiden name (i.e. Surname before marriage): Your Date of Birth: Date Month Year
Thank you for completing this questionnaire. Please return this questionnaire in
the reply-paid envelope supplied.
446
ID Number: __ __ __ - __ Staff Use Only
Centre for Accident Research and Road Safety – Queensland
Drivers 60 + Years Post-Test Questionnaire
August 2006
Associate Professor Jeremy Davey Deputy Director Centre for Accident Research & Road Safety – Queensland Queensland University of Technology Telephone: 07 3864 4574 Email: [email protected]
Angela Wallace PhD Scholar Centre for Accident Research & Road Safety – Queensland Queensland University of Technology Telephone: 0402 240 234 Email: [email protected]
447
Driving Self-Assessment
The next few questions relate to your driving. We would like for you to answer these questions as truthfully as possible. (Circle one number for each statement) N
ot a
t al
l
Very
O
ften
Are your reactions to unexpected situations slower than they used to be?
0 1 2 3 4 5
Do you have trouble judging the distance of other vehicles, or changing focus from your instrument panel to the road?
0 1 2 3 4 5
Are you having more trouble adjusting to glare and/or night driving than you did previously?
0 1 2 3 4 5
Do you ever get surprised by pedestrians or other vehicles coming from your left or right while you are focusing straight ahead?
0 1 2 3 4 5
Do some traffic situations or other drivers upset you?
0 1 2 3 4 5
Do you have trouble driving through, or turning at busy intersections or roundabouts?
0 1 2 3 4 5
Do you feel uncomfortable driving in unfamiliar territory?
0 1 2 3 4 5
Do you find that you are easily distracted or that your thoughts wander while you are driving?
0 1 2 3 4 5
Do you have regular health and vision checks?
0 1 2 3 4 5
General Driving Behaviour No one is perfect. Even the best drivers make mistakes, do foolish things, or bend the rules at some time or another. Some of these behaviours are trivial, but some are potentially dangerous. For each item below you are asked to indicate HOW OFTEN, if at all, this kind of thing has happened to you. On the scale of 0 ‘Never’ to 5 ‘Nearly all the time’, please circle the number which best sums up your answer.
(Circle one number for each statement) How often do you…..
Nev
er
Har
dly
ever
Occ
asio
nally
Qui
te o
ften
Freq
uent
ly
Nea
rly a
ll th
e tim
e
Attempt to overtake someone that you hadn’t noticed to be signalling a right turn
0 1 2 3 4 5
Get into the wrong lane when approaching a roundabout or a junction
0 1 2 3 4 5
Miss ‘Stop’ or ‘Give Way’ signs and narrowly avoid colliding with traffic having right of way
0 1 2 3 4 5
448
(Circle one number for each statement) How often do you…..
Nev
er
Har
dly
ever
Occ
asio
nally
Qui
te o
ften
Freq
uent
ly
Nea
rly a
ll th
e tim
e
Misread the signs and exit from the roundabout on the wrong road
0 1 2 3 4 5
Fail to notice that pedestrians are crossing when turning into a side street from a main road
0 1 2 3 4 5
Drive especially close to the car in front as a signal to its driver to go faster or get out of the way
0 1 2 3 4 5
Forget where you left your car in the car park
0 1 2 3 4 5
Queuing to turn left onto a main road, you pay such close attention to the mainstream of traffic that you nearly hit the car in front
0 1 2 3 4 5
Hit something when reversing that you had not previously seen 0 1 2 3 4 5
Cross a junction knowing that the traffic lights have already turned against you
0 1 2 3 4 5
On turning left nearly hit a cyclist who has come up on your inside 0 1 2 3 4 5
Disregard the speed limits late at night or very early in the morning 0 1 2 3 4 5
Attempt to drive away from the traffic lights in third gear 0 1 2 3 4 5
Fail to check your rear-view mirror before pulling out, changing lanes, etc.
0 1 2 3 4 5
Have an aversion to a particular class of road user, and indicate your hostility by whatever means you can
0 1 2 3 4 5
Become impatient with a slow driver in the outer lane and overtake on the inside
0 1 2 3 4 5
Underestimate the speed of an oncoming vehicle when overtaking 0 1 2 3 4 5
Switch on one thing, such as the headlights, when you meant to switch on something else, such as the wipers
0 1 2 3 4 5
Brake too quickly on a slippery road, or steer the wrong way in a skid 0 1 2 3 4 5
Intending to drive to destination A, you ‘wake up’ to find yourself on the road to destination B, perhaps because the latter is your more usual destination
0 1 2 3 4 5
Drive even though you realise you may be over the legal blood-alcohol limit
0 1 2 3 4 5
Get involved in unofficial ‘races’ with other drivers 0 1 2 3 4 5
Realise that you have no clear recollection of the road along which you have just been travelling
0 1 2 3 4 5
Angered by another driver’s behaviour, you give chase with the intention of giving him/her a piece of your mind
0 1 2 3 4 5
449
Driving at Railway Level Crossings During the past month, how often have you ever engaged in the following behaviours at a railway level crossing? On the scale of 0 ‘Not at all’ to 5 ‘Very often’, please circle the number which best sums up your answer. If you have not driven through any type of railway level crossing during the past month, (i.e. with or without boom gates), please circle ‘Not Applicable’.
(Circle one number for each statement) I have: N
ot a
t all
Very
O
ften
Not
App
licab
le
Driven through a rail crossing when lights are flashing, but boom gates are yet to descend
0 1 2 3 4 5 n/a
Driven through a rail crossing when boom gates are descending
0 1 2 3 4 5 n/a
Driven through a rail crossing when booms gates are beginning to rise, but lights are still flashing
0 1 2 3 4 5 n/a
Rolling through a rail crossing without stopping, if no train is visible
0 1 2 3 4 5 n/a
Driven through a rail crossing after the first train has passed, without looking for a second
0 1 2 3 4 5 n/a
Failed to look for a train before crossing a rail crossing
0 1 2 3 4 5 n/a
Driven through a rail crossing when the train is visible, but still some distance away
0 1 2 3 4 5 n/a
Driven through a rail crossing when the train is close
0 1 2 3 4 5 n/a
Queued over a rail crossing
0 1 2 3 4 5 n/a
Driven around boom gates to cross a rail crossing
0 1 2 3 4 5 n/a
Tried to beat the train across the rail crossing
0 1 2 3 4 5 n/a
Sped on approach to a rail crossing
0 1 2 3 4 5 n/a
Driven through a rail crossing when the lights are flashing, but no train is visible
0 1 2 3 4 5 n/a
Driven through a rail crossing when the lights are flashing, and the train is visible
0 1 2 3 4 5 n/a
Scanned on approach to a rail crossing, but failed to stop at it
0 1 2 3 4 5 n/a
Stopped on the yellow hatching road markings
0 1 2 3 4 5 n/a
Followed another vehicle across the rail crossing without looking for myself
0 1 2 3 4 5 n/a
Driven through a rail crossing when visibility is impaired
0 1 2 3 4 5 n/a
Driven through a rail crossing without realising it until after the vehicle is over the tracks
0 1 2 3 4 5 n/a
Seen police enforcing the road rules at a rail crossing
0 1 2 3 4 5 n/a
450
We would like to know the likelihood that you will engage in the following behaviours in the next 6 months of driving at railway level crossings. Even if you rarely drive through railway level crossings, please indicate how you would drive if you were to drive through a crossing. On the scale of 0 ‘Not at all likely’ to 5 ‘Very likely’, please circle the number which best describes your answer. (Circle one number for each statement) In the next 6 months of driving, it is likely I will: N
ot a
t all
likel
y
Very
lik
ely
Drive through a rail crossing when lights are flashing, but boom gates are yet to descend
0 1 2 3 4 5
Drive through a rail crossing when boom gates are descending
0 1 2 3 4 5
Drive through a rail crossing when booms gates are beginning to rise, but lights are still flashing
0 1 2 3 4 5
Roll through a rail crossing without stopping, if no train is visible
0 1 2 3 4 5
Drive through a rail crossing after the first train has passed, without looking for a second
0 1 2 3 4 5
Fail to look for a train before crossing a rail crossing
0 1 2 3 4 5
Drive through a rail crossing when the train is visible, but still some distance away
0 1 2 3 4 5
Drive through a rail crossing when the train is close
0 1 2 3 4 5
Queue over a rail crossing
0 1 2 3 4 5
Drive around boom gates to cross
0 1 2 3 4 5
Try to beat the train across the rail crossing
0 1 2 3 4 5
Speed on approach to a rail crossing
0 1 2 3 4 5
Drive through a rail crossing when the lights are flashing, but no train is visible
0 1 2 3 4 5
Drive through a rail crossing when the lights are flashing, and the train is visible
0 1 2 3 4 5
Scan on approach to a rail crossing, but fail to stop at it
0 1 2 3 4 5
Stop on the yellow hatching road markings
0 1 2 3 4 5
Follow another vehicle across the rail crossing without looking for myself
0 1 2 3 4 5
Drive through a rail crossing when visibility is impaired
0 1 2 3 4 5
Drive through a rail crossing without realising until after the vehicle is over the tracks
0 1 2 3 4 5
See police enforcing the road rules at a rail crossing
0 1 2 3 4 5
451
Rate each of the following statements according to the degree they reflect your personal views. On the scale of ‘1’ ‘Strongly Disagree’ to ‘6’ ‘Strongly Agree’, please circle the number which best describes your answer. (Circle one number for each statement)
Stro
ngly
D
isag
ree
Stro
ngly
A
gree
Your family generally obeys the rules at rail crossings 1 2 3 4 5 6
Your friends generally obey the rules at rail crossings 1 2 3 4 5 6
Other motorists generally obey the rules at rail crossings 1 2 3 4 5 6
Your family generally think it important to obey the rules at rail crossings 1 2 3 4 5 6
Your friends generally think it important to obey the rules at rail crossings 1 2 3 4 5 6
Other motorists generally think it important to obey the rules at rail crossings
1 2 3 4 5 6
It is generally safe to disobey the rules at rail crossings 1 2 3 4 5 6
It is generally possible to judge a train’s speed 1 2 3 4 5 6
It is generally safe to cross if you can’t see a train, even if the lights are flashing
1 2 3 4 5 6
It is generally safe to roll slowly through a crossing instead of stopping 1 2 3 4 5 6
Trains generally run to a regular timetable 1 2 3 4 5 6
Penalties need to be tougher for violating road rules at rail crossings 1 2 3 4 5 6
The main deterrent for breaking the rules at rail crossings is fear of getting caught
1 2 3 4 5 6
Generally it is more important to use common sense at rail crossings than strictly follow the road rules
1 2 3 4 5 6
We would like to ask you some questions about your attitude towards railway level crossings. Please circle one number in each table. As a motorist, I believe that the design of rail crossings are:
Bad Good -3 -2 -1 0 1 2 3
Unsafe Safe
-3 -2 -1 0 1 2 3
As a motorist, I believe that the design of rail crossings are:
Confusing Easily Understood
-3 -2 -1 0 1 2 3
452
As a motorist, I believe that the design of rail crossings are:
Difficult for obeying
road rules
Easy for obeying
road rules -3 -2 -1 0 1 2 3
As a motorist, I believe that road rules at rail crossings are:
Bad
Good
-3 -2 -1 0 1 2 3
Not Strict Enough
Too Strict
-3 -2 -1 0 1 2 3
As a motorist, I believe that road rules at rail crossings are:
Confusing
Easily Understood
-3 -2 -1 0 1 2 3
Not
Practical
Practical -3 -2 -1 0 1 2 3
Obeying the road rules at rail crossings is:
Not up to me
Up to me
-3 -2 -1 0 1 2 3
Out of my
control Under my
control -3 -2 -1 0 1 2 3
Obeying the road rules at rail crossings is:
Dependent on other
motorists
Dependent on me only
-3 -2 -1 0 1 2 3
453
Obeying the road rules at rail crossings is:
Dependent
on time constraints
Not dependent
on time constraints
-3 -2 -1 0 1 2 3
Other motorists influence my driving behaviour at rail crossings, making it:
Harder to obey road
rules
Easier to obey road
rules -3 -2 -1 0 1 2 3
More
confusing Less
confusing -3 -2 -1 0 1 2 3
More
stressful More
relaxing -3 -2 -1 0 1 2 3
When you are driving a car at railway level crossings, how often do each of the following occur? On the scale of 0 ‘Never’ to ‘5’ ‘Always’, please circle the number which best describes your answer.
(Circle one number for each statement) N
ever
Alw
ays
Blinding sun makes it difficult to see if the red flashing lights are activated
0 1 2 3 4 5
The design of the road makes it difficult to see if a train is approaching or at the rail crossing
0 1 2 3 4 5
Warning systems on the road approaching the crossing are not adequate to inform drivers there is a rail crossing ahead
0 1 2 3 4 5
Road surfaces are often poor and it is difficult to stop
0 1 2 3 4 5
Boom gates and/or flashing lights are often faulty
0 1 2 3 4 5
Other drivers do stupid things that put you in a dangerous situation
0 1 2 3 4 5
Intersections ahead of a rail crossing often cause your car to overhang the tracks
0 1 2 3 4 5
Difficult to hear an approaching train when the windows are up 0 1 2 3 4 5 My car stalled on the tracks at a rail crossing
0 1 2 3 4 5
454
The following section tests your knowledge about railway level crossings. Circle the number that best describes your answer. When the twin red lights start flashing this means?
A train is coming but it is still safe to cross………….. You must stop and not enter the crossing ………….. Don’t know ………………………………………………
1 2 3
Boom gates usually stay down a lot longer than traffic lights take to change at an intersection
True ………………………………………………………False ……………........…………………………………. Don’t know ………………………………………………
1 2 3
Yellow hatching road markings means keep clear
True ………………………………………………………False …………………………………………………….. Don’t know ………………………………………………
1 2 3
There are fines for not stopping at railway crossings
True ………………………………………………………False …………………………………………………….. Don’t know ………………………………………………
1 2 3
How likely is it, that as a Driver you will be involved in a crash with a train at a railway level crossing?
Not at all Likely
Very Likely
0 1 2 3 4 5
Road Safety Radio Message We would now like to ask you some questions about the road safety advertisement that you listened to recently for this project when telephoned by one of our researchers.
Which road safety message did you listen to on the telephone?
Vision and driving ………...……………………….…… Driving at railway level crossing …...…………………
1 2
Can you remember any slogans from the road safety advertisement that you listened to on the telephone? If YES, which slogan do you recall?
No ..………………………………………………….…… Yes .....…………………………………………………… Slogan: ___________________________________________
1 2
Can you remember any important information mentioned in the advertisement? If YES, what important information do you recall?
No ..………………………………………………….…… Yes .....…………………………………………………… Important information: ___________________________________________
1 2
455
How likely is it, that your driving behaviour has been affected by listening to this road safety advertisement on the telephone?
Not at all
likely Very likely
0 1 2 3 4 5
Thank you for completing this questionnaire. Please return this questionnaire in
the reply-paid envelope supplied.
456
Company: ______________ (Staff Use Only)
ID Number: ____ ____ ____
Centre for Accident Research and Road Safety – Queensland
Heavy Vehicle Drivers Pre-Test Questionnaire
July 2006
Associate Professor Jeremy Davey Deputy Director Centre for Accident Research & Road Safety – Queensland Queensland University of Technology Telephone: 07 3864 4574 Email: [email protected]
Angela Wallace PhD Scholar Centre for Accident Research & Road Safety – Queensland Queensland University of Technology Telephone: 0402 240 234 Email: [email protected]
457
1. Personal Information We would like to start by asking you some questions about yourself. Please circle the number that best describes you. a) What is your gender? Male………………………………………………….……
Female…………………………………………………… 1 2
b) What is your age? _________ Years
c) What is your employment status? Full-time ………….. ...………………………..…………
Part-time / Casual ……… …………………..….……… Non-employee (agency, sub-contractor or haulier) …
1 2 3
d) What type of employment do you have with the company?
Company employee ………….. ...…………………..… Sub-contractor ………………………….……............... Agency employee ………………………………………
1 2 3
e) Do you work shift work? If YES – please circle your working hours.
Yes ………………………………………………………. No ………………………………………………………... Night and shift work …………………….....……...…… Shift work, no night work …..……………..…………… Night work, no shift work ….....………………………..
1 2
1 2 3
f) What is the postcode where you live?
g) What type of Licence do you hold? Class HR (Heavy Rigid Truck) ………….. ...………… Class HC (Heavy Combination) …………...…………. Class MC (B-double, Road Train) ……......................
1 2 3
h) How many years have you held your Heavy Vehicle Licence for?
_________ Years
2. Driving Patterns We would like to ask you some questions about how often you drive a Truck. a) How often do you drive a Truck, assuming an average week?
Every day of the week ………………………………… 4 – 6 days a week.....………….………………………. 2 – 3 days a week..……………………………………. At least 1 day a week .………………………………… Less than 1 day a week ……….……………………....
1 2 3 4 5
b) How many hours do you drive a Truck each week?
1 – 40 hrs ……………………………….....……....…… 41 – 50 hrs……. …………….……..…………………… 51 – 60 hrs.……..…...………………….………………. 61 – 72 hrs ..………………………………………..…… More than 72 hrs ………………………………………..
1 2 3 4 5
458
c) How many kilometres do you drive a Truck in an average week?
0 – 999 ………....…………..…………………………… 1000 – 1999….…………………………………………. 2000 – 2999 …..…….………………………………….. 3000 – 3999 ……..……….…………………………….. 4000 – 4999 ……………………………………………. 5000 or more .…….……………………………………..
1 2 3 4 5 6
3. Road Crashes We would now like to ask you some questions about any road crashes you have been involved in as Truck Driver in the past 3 years. By crash, we mean any collision that involved either injury to another person or yourself, damage to property, damage to another vehicle, or damage to the vehicle you were driving. Please mention only those crashes that you were involved in whilst driving a Truck. a) How many crashes have you been involved in as a Truck Driver in the past 3 years?
__________
If you have not been in a crash as a Truck Driver in the past 3 years, please skip to Question 4. b) Of the crashes that you have been involved in as a Truck Driver, how many resulted in: Damage only ______________ Slight injury (to any person)
______________
Serious injury (to any person)
______________
Fatality (of any person)
______________
c) Did any of the road crashes occur at a railway level crossing? Yes ……………………………………...… No ……………………………………........
1 2
d) Thinking about your most recent crash whilst driving a Truck, what type of collision was it?
(Circle one number on each line)
Yes No
Head on collision with another vehicle
1 0
Rear-end collision with another vehicle
1 0
Angular (i.e. side-on) collision with another vehicle
1 0
Collision with a pedestrian
1 0
Collision with another object (e.g. parked car, animal, tree)
1 0
Overturned vehicle
1 0
Other 1 0
459
e) Was the crash your fault?
Not at all Partially Fully (Circle one number only) 0 1 2
f) How true were the following statements at the time of the crash? (Circle one number on each line) N
ot T
rue
Som
ewha
t Tru
e
Very
Tru
e
I had been drinking alcohol before driving
0 1 2
I felt tired
0 1 2
I was driving too fast for the conditions
0 1 2
I was talking to a passenger in the vehicle
0 1 2
I was listening to music or the CB radio
0 1 2
I was talking on a mobile phone
0 1 2
I was trying to pick up something from the seat or floor
0 1 2
I was adjusting the radio/cassette/CD or fan/air conditioning
0 1 2
I was checking the instruments (e.g. fuel gauge or speedometer)
0 1 2
I was daydreaming
0 1 2
I was talking on the CB radio
0 1 2
I was distracted by something inside the vehicle
0 1 2
I was distracted by something outside the vehicle
0 1 2
I couldn’t see the other vehicle / object
0 1 2
4. General Driving Behaviour The following are statements that may describe your driving behaviour whilst driving a Truck. On the scale of ‘1’ ‘Never’ to ‘6’ ‘Nearly all the time’ please circle the number which best sums up your answer.
(Circle one number on each line) How often do you…..
Nev
er
Har
dly
ever
Occ
asio
nally
Qui
te o
ften
Freq
uent
ly
Nea
rly a
ll th
e tim
e
Attempt to overtake someone in front of you that you hadn’t noticed to be turning in front of you
1 2 3 4 5 6
Stay in a lane that you know will be closed ahead until the last minute before forcing your way into another lane
1 2 3 4 5 6
460
(Circle one number on each line) How often do you…..
Nev
er
Har
dly
ever
Occ
asio
nally
Qui
te o
ften
Freq
uent
ly
Nea
rly a
ll th
e tim
e
Miss ‘Stop’ or ‘Give Way’ signs 1 2 3 4 5 6
Intentionally disobey a ‘Stop’ or ‘Give Way’ sign 1 2 3 4 5 6
Pull out of a junction so far that you disrupt the flow of traffic 1 2 3 4 5 6
Fail to notice that pedestrians are crossing in your path of traffic 1 2 3 4 5 6
Drive especially close to the car in front as a signal to its driver to go faster or get out of the way
1 2 3 4 5 6
Sound your horn to indicate your annoyance to another driver 1 2 3 4 5 6
Queuing to enter a main road, you pay such close attention to the mainstream of traffic that you nearly hit the car in front
1 2 3 4 5 6
Cross a junction knowing that the traffic lights have already turned against you
1 2 3 4 5 6
Whilst turning nearly hit a cyclist who has come up on your inside 1 2 3 4 5 6
Intentionally disregard the speed limit on a highway/freeway 1 2 3 4 5 6
Exceed the speed limit on a highway/freeway without realising 1 2 3 4 5 6
Fail to check your rear-view mirror before pulling out, changing lanes, etc
1 2 3 4 5 6
Become angered by a certain type of driver and indicate your hostility by whatever means you can
1 2 3 4 5 6
Become impatient with a slow driver ahead and overtake on the inside
1 2 3 4 5 6
When overtaking underestimate the speed of an oncoming vehicle 1 2 3 4 5 6
Race away from the traffic lights with the intention of beating the driver next to you
1 2 3 4 5 6
Skid while braking or cornering on a slippery road 1 2 3 4 5 6
Drive even though you suspect you may be over the legal blood-alcohol limit
1
2
3
4
5
6
Intentionally disregard the speed limit on a residential road 1 2 3 4 5 6
Exceed the speed limit on a residential road without realising 1 2 3 4 5 6
Become angered by another driver and give chase 1 2 3 4 5 6
Hit something while manoeuvring (including parking and reversing) 1 2 3 4 5 6
461
(Circle one number on each line) How often do you…..
Nev
er
Har
dly
ever
Occ
asio
nally
Qui
te o
ften
Freq
uent
ly
Nea
rly a
ll th
e tim
e
Come close to hitting something while manoeuvring (including parking and reversing)
1 2 3 4 5 6
Drive while under time pressure 1 2 3 4 5 6
Find your attention being distracted from the road 1 2 3 4 5 6
Drive while tired 1 2 3 4 5 6
Have difficulty driving because of tiredness or fatigue 1 2 3 4 5 6
Find yourself nodding off while driving for work 1 2 3 4 5 6
Lose concentration while driving 1 2 3 4 5 6
Not wear your seatbelt 1 2 3 4 5 6
Remove your seatbelt for some reason while driving 1 2 3 4 5 6
Drive while using a ‘hand-held’ mobile phone 1 2 3 4 5 6
Drive while using a ‘hands-free’ mobile phone 1 2 3 4 5 6
Check the tyre pressure and fluid levels of your vehicle 1 2 3 4 5 6
Do paperwork or other administration whilst driving 1 2 3 4 5 6
Save the time during the day by driving quicker between places 1 2 3 4 5 6
Have one or two alcoholic drinks before driving 1 2 3 4 5 6
Smoke a cigarette whilst driving 1 2 3 4 5 6
Eat food whilst driving 1 2 3 4 5 6
5. Driving at Railway Level Crossings a) Please circle how frequently you go through the following types of railway crossing whilst driving a Truck.
(Please only circle one number for each statement)
Dai
ly
Wee
kly
Mon
thly
Year
ly
Nev
er
Boom gates with flashing lights
1 2 3 4 0
Flashing lights only (no Boom Gate)
1 2 3 4 0
Only STOP or GIVE WAY sign
1 2 3 4 0
462
b) How often have you engaged in the following behaviours whilst driving a Truck at a railway level crossing? On the scale of ‘1’ ‘Not at all’ to ‘5’ ‘Very often’, please circle the number which best sums up how much you agree with the following statements
(Circle one number on each line)
Not
at a
ll
Very
Ofte
n
Driven through a crossing when lights are flashing, but boom gates are yet to descend
1 2 3 4 5
Driven through a crossing when boom gates are descending
1 2 3 4 5
Driven through a crossing when booms gates are begin to rise, but lights are still flashing
1 2 3 4 5
Rolling through a crossing without stopping, if no train is visible
1 2 3 4 5
Driven through a crossing after the first train has passed, without looking for a second
1 2 3 4 5
Failed to look for a train before crossing
1 2 3 4 5
Driven through a crossing when the train is visible, but still some distance away
1 2 3 4 5
Driven through a crossing when the train is close
1 2 3 4 5
Queued over a crossing
1 2 3 4 5
Driven around boom gates to cross
1 2 3 4 5
Tried to beat the train across the crossing
1 2 3 4 5
Sped on approach to a crossing
1 2 3 4 5
Scanned on approach to a crossing, but failed to stop at it
1 2 3 4 5
Driven through a crossing when the lights are flashing, but no train is visible
1 2 3 4 5
Driven through a crossing when the lights are flashing, and the train is visible
1 2 3 4 5
Stopped on the yellow hatching road markings
1 2 3 4 5
Followed another vehicle across the crossing without looking for yourself
1 2 3 4 5
Driven through a crossing when visibility is impaired 1 2 3 4 5
Driven through a crossing without realising it until after the vehicle is over the tracks
1 2 3 4 5
463
c) We would like to know the likelihood that you will engage in the following behaviours in the next 6 months of driving at railway level crossings. Even if you rarely drive through railway level crossings, please indicate how you would drive if you were to drive through a crossing. On the scale of ‘1’ ‘Not at all’ to ‘5’ ‘Very often’, please circle the number which best sums up how much you agree with the following statements.
(Circle one number on each line) In the next 6 months of driving, it is likely I will:
Not
at a
ll Li
kely
Very
Lik
ely
Drive through a crossing when lights are flashing, but boom gates are yet to descend
1 2 3 4 5
Drive through a crossing when boom gates are descending
1 2 3 4 5
Drive through a crossing when booms gates are begin to rise, but lights are still flashing
1 2 3 4 5
Roll through a crossing without stopping, if no train is visible
1 2 3 4 5
Drive through a crossing after the first train has passed, without looking for a second
1 2 3 4 5
Fail to look for a train before crossing
1 2 3 4 5
Drive through a crossing when the train is visible, but still some distance away
1 2 3 4 5
Drive through a crossing when the train is close
1 2 3 4 5
Queue over a crossing
1 2 3 4 5
Drive around boom gates to cross
1 2 3 4 5
Try to beat the train across the crossing
1 2 3 4 5
Speed on approach to a crossing
1 2 3 4 5
Scan on approach to a crossing, but failed to stop at it
1 2 3 4 5
Drive through a crossing when the lights are flashing, but no train is visible
1 2 3 4 5
Drive through a crossing when the lights are flashing, and the train is visible
1 2 3 4 5
Stop on the yellow hatching road markings
1 2 3 4 5
Follow another vehicle across the crossing without looking for yourself
1 2 3 4 5
Drive through a crossing when visibility is impaired 1 2 3 4 5
Drive through a crossing without realising until after the vehicle is over the tracks
1 2 3 4 5
464
d) Rate each of the following statements according to the degree they reflect your personal opinion. On the scale of ‘1’ ‘Not at all’ to ‘5’ ‘Very often’, please circle the number which best sums up how much you agree or disagree with the following statements. (Please only circle one number for each statement)
Stro
ngly
A
gree
Stro
ngly
D
isag
ree
Your colleagues generally obey the rules at level crossings
1 2 3 4 5
Your friends generally obey the rules at level crossings
1 2 3 4 5
Other motorists generally obey the rules at level crossings
1 2 3 4 5
Your colleagues generally think it important to obey the rules at level crossings
1 2 3 4 5
Your friends generally think it important to obey the rules at level crossings
1 2 3 4 5
Other motorists generally think it important to obey the rules at level crossings
1 2 3 4 5
It is generally safe to disobey the rules at level crossings
1 2 3 4 5
It is generally possible to judge a train’s speed
1 2 3 4 5
It is generally safe to cross if you can’t see a train, even if the lights are flashing
1 2 3 4 5
It is generally safe to roll slowly through a crossing instead of stopping
1 2 3 4 5
Trains generally run to a regular timetable
1 2 3 4 5
Penalties need to be tougher for violating road rules at level crossings
1 2 3 4 5
The main deterrent for breaking the rules at level crossings is fear of getting caught
1 2 3 4 5
Generally it is more important to use common sense at level crossings than strictly follow the road rules
1 2 3 4 5
Your colleagues generally obey the road rules at level crossings
1 2 3 4 5
Your friends generally obey the rules at level crossings
1 2 3 4 5
Other motorists generally obey the rules at level crossings
1 2 3 4 5
Your colleagues generally think it important to obey the rules at level crossings
1 2 3 4 5
465
e) We would like to ask you some questions about your attitude towards railway level crossings. Please circle one number in each table. As a truck driver, I believe that the design of railway level crossings is:
Bad Good
-3 -2 -1 0 1 2 3
Unsafe Safe
-3 -2 -1 0 1 2 3
Confusing Easily
Understood -3 -2 -1 0 1 2 3
Difficult for
obeying road rules
Easy for obeying
road rules -3 -2 -1 0 1 2 3
As a truck driver, I believe that road rules at railway level crossings are:
Bad Good
-3 -2 -1 0 1 2 3
Not Strict Enough
Too Strict
-3 -2 -1 0 1 2 3
Confusing Easily
Understood -3 -2 -1 0 1 2 3
Not
Practical Practical
-3 -2 -1 0 1 2 3
Obeying the road rules at railway level crossings is:
Not up to me
Up to me
-3 -2 -1 0 1 2 3
Out of my
control Under my
control -3 -2 -1 0 1 2 3
466
Obeying the road rules at railway level crossings is:
Dependent on other
motorists
Dependent on me only
-3 -2 -1 0 1 2 3
Dependent
on time constraints
Not dependent
on time constraints
-3 -2 -1 0 1 2 3
Other motorists influence my ability to obey road rules at railway level crossings, making it:
Harder Easier -3 -2 -1 0 1 2 3
More
confusing Less
confusing -3 -2 -1 0 1 2 3
More
stressful More
relaxing -3 -2 -1 0 1 2 3
f) When you are driving a Truck at railway level crossings, how often do each of the following occur?
(Please only circle one number for each statement) N
ever
Alm
ost N
ever
Som
etim
es
Alm
ost A
lway
s
Alw
ays
Blinding sun makes it difficult to see if the red flashing lights are activated
1 2 3 4 5
The design of the road is an ‘S’ bend and it is difficult to see if a train is approaching or at the railway crossing
1 2 3 4 5
The height of the truck’s cabin makes it difficult to see a train or the warning systems
1 2 3 4 5
Warning systems on the road approaching the crossing are not adequate to inform trucks there is a railway crossing ahead
1 2 3 4 5
The mass of the truck makes it difficult to brake in time to stop at a railway crossing
1 2 3 4 5
Road surface is often poor and it is difficult to stop
1 2 3 4 5
Boom gates and/or flashing lights are often faulty
1 2 3 4 5
467
(Please only circle one number for each statement) N
ever
Alm
ost N
ever
Som
etim
es
Alm
ost A
lway
s
Alw
ays
Other drivers (such as cars) do stupid things that put you in a dangerous situation
1 2 3 4 5
Intersections ahead of a railway crossing often cause your truck to overhang the tracks
1 2 3 4 5
Noise from the truck’s engine is too loud to hear an approaching train
1 2 3 4 5
When you have to stop at a railway crossing, it takes a long time before your truck is able to get over the crossing
1 2 3 4 5
All ‘Stop’ signs should be changed to ‘Give Way’ signs at railway crossings as these are easier for trucks to get through the crossing
1 2 3 4 5
My truck has stalled on the tracks at a railway crossing 1 2 3 4 5 g) The following section tests your knowledge about railway level crossings. Circle the number that best describes your answer. When the twin red lights start flashing this means?
A train is coming but it is still safe to cross………….. You must stop and not enter the crossing ………….. Don’t know ………………………………………………
1 2 3
Boom gates usually stay down a lot longer than traffic lights take to change at an intersection
True ………………………………………………………False ……………........…………………………………. Don’t know ………………………………………………
1 2 3
Yellow hatching road markings means keep clear
True ………………………………………………………False …………………………………………………….. Don’t know ………………………………………………
1 2 3
There are fines for not stopping at railway crossings
True ………………………………………………………False …………………………………………………….. Don’t know ………………………………………………
1 2 3
h) How likely is it that whilst driving a Truck that you will be involved in a crash with a train at a railway level crossing?
Not at all Likely
Very Likely
1 2 3 4 5
468
Finally, we would like to be able to match this questionnaire with others at a later date. However, we wish to preserve your anonymity. One way to do this is to collect some information from you that will not identify you, but will allow us to match the next questionnaire. Please assist us by providing: The month of your birth: __________ (eg. May) The year of your birth: __________ (eg. 1966) First 3 letters of your surname: _______________ (eg. Smi for Smith)
Thank you for completing this questionnaire.
469
Company: ______________ (Staff Use Only)
ID Number: ____ ____ ____
Centre for Accident Research and Road Safety – Queensland
Heavy Vehicle Drivers Post-Test Questionnaire
August 2006
Associate Professor Jeremy Davey Deputy Director Centre for Accident Research & Road Safety – Queensland Queensland University of Technology Telephone: 07 3864 4574 Email: [email protected]
Angela Wallace PhD Scholar Centre for Accident Research & Road Safety – Queensland Queensland University of Technology Telephone: 0402 240 234 Email: [email protected]
General Driving Behaviour
470
The following are statements that may describe your driving behaviour whilst driving a Truck. On the scale of ‘1’ ‘Never’ to ‘6’ ‘Nearly all the time’ please circle the number which best sums up your answer.
(Circle one number on each line) How often do you…..
Nev
er
Har
dly
ever
Occ
asio
nally
Qui
te o
ften
Freq
uent
ly
Nea
rly a
ll th
e tim
e
Attempt to overtake someone in front of you that you hadn’t noticed to be turning in front of you
1 2 3 4 5 6
Stay in a lane that you know will be closed ahead until the last minute before forcing your way into another lane
1 2 3 4 5 6
Miss ‘Stop’ or ‘Give Way’ signs 1 2 3 4 5 6
Intentionally disobey a ‘Stop’ or ‘Give Way’ sign 1 2 3 4 5 6
Pull out of a junction so far that you disrupt the flow of traffic 1 2 3 4 5 6
Fail to notice that pedestrians are crossing in your path of traffic 1 2 3 4 5 6
Drive especially close to the car in front as a signal to its driver to go faster or get out of the way
1 2 3 4 5 6
Sound your horn to indicate your annoyance to another driver 1 2 3 4 5 6
Queuing to enter a main road, you pay such close attention to the mainstream of traffic that you nearly hit the car in front
1 2 3 4 5 6
Cross a junction knowing that the traffic lights have already turned against you
1 2 3 4 5 6
Whilst turning nearly hit a cyclist who has come up on your inside 1 2 3 4 5 6
Intentionally disregard the speed limit on a highway/freeway 1 2 3 4 5 6
Exceed the speed limit on a highway/freeway without realising 1 2 3 4 5 6
Fail to check your rear-view mirror before pulling out, changing lanes, etc
1 2 3 4 5 6
Become angered by a certain type of driver and indicate your hostility by whatever means you can
1 2 3 4 5 6
Become impatient with a slow driver ahead and overtake on the inside
1 2 3 4 5 6
When overtaking underestimate the speed of an oncoming vehicle 1 2 3 4 5 6
Race away from the traffic lights with the intention of beating the driver next to you
1 2 3 4 5 6
Skid while braking or cornering on a slippery road 1 2 3 4 5 6
471
(Circle one number on each line) How often do you…..
Nev
er
Har
dly
ever
Occ
asio
nally
Qui
te o
ften
Freq
uent
ly
Nea
rly a
ll th
e tim
e
Drive even though you suspect you may be over the legal blood-alcohol limit
1
2
3
4
5
6
Intentionally disregard the speed limit on a residential road 1 2 3 4 5 6
Exceed the speed limit on a residential road without realising 1 2 3 4 5 6
Become angered by another driver and give chase 1 2 3 4 5 6
Hit something while manoeuvring (including parking and reversing) 1 2 3 4 5 6
Come close to hitting something while manoeuvring (including parking and reversing)
1 2 3 4 5 6
Drive while under time pressure 1 2 3 4 5 6
Find your attention being distracted from the road 1 2 3 4 5 6
Drive while tired 1 2 3 4 5 6
Have difficulty driving because of tiredness or fatigue 1 2 3 4 5 6
Find yourself nodding off while driving for work 1 2 3 4 5 6
Lose concentration while driving 1 2 3 4 5 6
Not wear your seatbelt 1 2 3 4 5 6
Remove your seatbelt for some reason while driving 1 2 3 4 5 6
Drive while using a ‘hand-held’ mobile phone 1 2 3 4 5 6
Drive while using a ‘hands-free’ mobile phone 1 2 3 4 5 6
Check the tyre pressure and fluid levels of your vehicle 1 2 3 4 5 6
Do paperwork or other administration whilst driving 1 2 3 4 5 6
Save the time during the day by driving quicker between places 1 2 3 4 5 6
Have one or two alcoholic drinks before driving 1 2 3 4 5 6
Smoke a cigarette whilst driving 1 2 3 4 5 6
Eat food whilst driving 1 2 3 4 5 6
472
Driving at Railway Level Crossings How often have you engaged in the following behaviours whilst driving a Truck at a railway level crossing in the past 2-3 weeks? On the scale of ‘1’ ‘Not at all’ to ‘5’ ‘Very often’, please circle the number which best sums up how much you agree with the following statements
(Circle one number on each line)
Not
at a
ll
Very
Ofte
n
Driven through a crossing when lights are flashing, but boom gates are yet to descend
1 2 3 4 5
Driven through a crossing when boom gates are descending
1 2 3 4 5
Driven through a crossing when booms gates are begin to rise, but lights are still flashing
1 2 3 4 5
Rolling through a crossing without stopping, if no train is visible
1 2 3 4 5
Driven through a crossing after the first train has passed, without looking for a second
1 2 3 4 5
Failed to look for a train before crossing
1 2 3 4 5
Driven through a crossing when the train is visible, but still some distance away
1 2 3 4 5
Driven through a crossing when the train is close
1 2 3 4 5
Queued over a crossing
1 2 3 4 5
Driven around boom gates to cross
1 2 3 4 5
Tried to beat the train across the crossing
1 2 3 4 5
Sped on approach to a crossing
1 2 3 4 5
Scanned on approach to a crossing, but failed to stop at it
1 2 3 4 5
Driven through a crossing when the lights are flashing, but no train is visible
1 2 3 4 5
Driven through a crossing when the lights are flashing, and the train is visible
1 2 3 4 5
Stopped on the yellow hatching road markings
1 2 3 4 5
Followed another vehicle across the crossing without looking for yourself
1 2 3 4 5
Driven through a crossing when visibility is impaired 1 2 3 4 5
Driven through a crossing without realising it until after the vehicle is over the tracks
1 2 3 4 5
473
We would like to know the likelihood that you will engage in the following behaviours in the next 6 months of driving at railway level crossings. Even if you rarely drive through railway level crossings, please indicate how you would drive if you were to drive through a crossing. On the scale of ‘1’ ‘Not at all’ to ‘5’ ‘Very often’, please circle the number which best sums up how much you agree with the following statements.
(Circle one number on each line) In the next 6 months of driving, it is likely I will:
Not
at a
ll Li
kely
Very
Lik
ely
Drive through a crossing when lights are flashing, but boom gates are yet to descend
1 2 3 4 5
Drive through a crossing when boom gates are descending
1 2 3 4 5
Drive through a crossing when booms gates are begin to rise, but lights are still flashing
1 2 3 4 5
Roll through a crossing without stopping, if no train is visible
1 2 3 4 5
Drive through a crossing after the first train has passed, without looking for a second
1 2 3 4 5
Fail to look for a train before crossing
1 2 3 4 5
Drive through a crossing when the train is visible, but still some distance away
1 2 3 4 5
Drive through a crossing when the train is close
1 2 3 4 5
Queue over a crossing
1 2 3 4 5
Drive around boom gates to cross
1 2 3 4 5
Try to beat the train across the crossing
1 2 3 4 5
Speed on approach to a crossing
1 2 3 4 5
Scan on approach to a crossing, but failed to stop at it
1 2 3 4 5
Drive through a crossing when the lights are flashing, but no train is visible
1 2 3 4 5
Drive through a crossing when the lights are flashing, and the train is visible
1 2 3 4 5
Stop on the yellow hatching road markings
1 2 3 4 5
Follow another vehicle across the crossing without looking for yourself
1 2 3 4 5
Drive through a crossing when visibility is impaired 1 2 3 4 5
Drive through a crossing without realising until after the vehicle is over the tracks
1 2 3 4 5
474
Rate each of the following statements according to the degree they reflect your personal opinion. On the scale of ‘1’ ‘Not at all’ to ‘5’ ‘Very often’, please circle the number which best sums up how much you agree or disagree with the following statements. (Please only circle one number for each statement)
Stro
ngly
A
gree
Stro
ngly
D
isag
ree
Your colleagues generally obey the rules at level crossings
1 2 3 4 5
Your friends generally obey the rules at level crossings
1 2 3 4 5
Other motorists generally obey the rules at level crossings
1 2 3 4 5
Your colleagues generally think it important to obey the rules at level crossings
1 2 3 4 5
Your friends generally think it important to obey the rules at level crossings
1 2 3 4 5
Other motorists generally think it important to obey the rules at level crossings
1 2 3 4 5
It is generally safe to disobey the rules at level crossings
1 2 3 4 5
It is generally possible to judge a train’s speed
1 2 3 4 5
It is generally safe to cross if you can’t see a train, even if the lights are flashing
1 2 3 4 5
It is generally safe to roll slowly through a crossing instead of stopping
1 2 3 4 5
Trains generally run to a regular timetable
1 2 3 4 5
Penalties need to be tougher for violating road rules at level crossings
1 2 3 4 5
The main deterrent for breaking the rules at level crossings is fear of getting caught
1 2 3 4 5
Generally it is more important to use common sense at level crossings than strictly follow the road rules
1 2 3 4 5
475
We would like to ask you some questions about your attitude towards railway level crossings. Please circle one number in each table. As a truck driver, I believe that the design of railway level crossings is:
Bad Good
-3 -2 -1 0 1 2 3
Unsafe Safe
-3 -2 -1 0 1 2 3
Confusing Easily
Understood -3 -2 -1 0 1 2 3
Difficult for
obeying road rules
Easy for obeying
road rules -3 -2 -1 0 1 2 3
As a truck driver, I believe that road rules at railway level crossings are:
Bad Good
-3 -2 -1 0 1 2 3
Not Strict Enough
Too Strict
-3 -2 -1 0 1 2 3
Confusing Easily
Understood -3 -2 -1 0 1 2 3
Not
Practical Practical
-3 -2 -1 0 1 2 3
Obeying the road rules at railway level crossings is:
Not up to me
Up to me
-3 -2 -1 0 1 2 3
Out of my
control Under my
control -3 -2 -1 0 1 2 3
476
Obeying the road rules at railway level crossings is:
Dependent on other
motorists
Dependent on me only
-3 -2 -1 0 1 2 3
Dependent
on time constraints
Not dependent
on time constraints
-3 -2 -1 0 1 2 3
Other motorists influence my ability to obey road rules at railway level crossings, making it:
Harder Easier -3 -2 -1 0 1 2 3
More
confusing Less
confusing -3 -2 -1 0 1 2 3
More
stressful More
relaxing -3 -2 -1 0 1 2 3
When you are driving a Truck at railway level crossings, how often do each of the following occur?
(Please only circle one number for each statement) N
ever
Alm
ost N
ever
Som
etim
es
Alm
ost A
lway
s
Alw
ays
Blinding sun makes it difficult to see if the red flashing lights are activated
1 2 3 4 5
The design of the road is an ‘S’ bend and it is difficult to see if a train is approaching or at the railway crossing
1 2 3 4 5
The height of the truck’s cabin makes it difficult to see a train or the warning systems
1 2 3 4 5
Warning systems on the road approaching the crossing are not adequate to inform trucks there is a railway crossing ahead
1 2 3 4 5
The mass of the truck makes it difficult to brake in time to stop at a railway crossing
1 2 3 4 5
Road surface is often poor and it is difficult to stop
1 2 3 4 5
Boom gates and/or flashing lights are often faulty 1 2 3 4 5
477
(Please only circle one number for each statement) N
ever
Alm
ost N
ever
Som
etim
es
Alm
ost A
lway
s
Alw
ays
Other drivers (such as cars) do stupid things that put you in a dangerous situation
1 2 3 4 5
Intersections ahead of a railway crossing often cause your truck to overhang the tracks
1 2 3 4 5
Noise from the truck’s engine is too loud to hear an approaching train
1 2 3 4 5
When you have to stop at a railway crossing, it takes a long time before your truck is able to get over the crossing
1 2 3 4 5
All ‘Stop’ signs should be changed to ‘Give Way’ signs at railway crossings as these are easier for trucks to get through the crossing
1 2 3 4 5
My truck has stalled on the tracks at a railway crossing 1 2 3 4 5 The following section tests your knowledge about railway level crossings. Circle the number that best describes your answer. When the twin red lights start flashing this means?
A train is coming but it is still safe to cross………….. You must stop and not enter the crossing ………….. Don’t know ………………………………………………
1 2 3
Boom gates usually stay down a lot longer than traffic lights take to change at an intersection
True ………………………………………………………False ……………........…………………………………. Don’t know ………………………………………………
1 2 3
Yellow hatching road markings means keep clear
True ………………………………………………………False …………………………………………………….. Don’t know ………………………………………………
1 2 3
There are fines for not stopping at railway crossings
True ………………………………………………………False …………………………………………………….. Don’t know ………………………………………………
1 2 3
How likely is it that whilst driving a Truck that you will be involved in a crash with a train at a railway level crossing?
Not at all Likely
Very Likely
1 2 3 4 5
478
Road Safety Radio Message We would now like to ask you to cast your mind back to the audio message played to you several months ago by one of our researchers. The following questions refer to this road safety message.
Which road safety message did you listen to on the telephone?
Fatigue and driving ………...……………………….……Driving at railway level crossings ...………………...…
1 2
Can you remember any slogans from the road safety advertisement that you listened to? If YES, which slogan do you recall?
No ..………………………………………………….…… Yes .....…………………………………………………… Slogan: ___________________________________________
1 2
Can you remember any important information mentioned in the advertisement? If YES, what important information do you recall?
No ..………………………………………………….…… Yes .....…………………………………………………… Important information: ___________________________________________
1 2
How likely is it, that your driving behaviour has been affected by listening to this road safety advertisement for heavy vehicle driving?
Not at all
likely Very likely
0 1 2 3 4 5
Thank you for completing this questionnaire. Please return this questionnaire in
the reply-paid envelope supplied.