beyond compliance n exploratory investigation of … · 2019-07-24 · beyond compliance: an...

BEYOND COMPLIANCE: AN

EXPLORATORY INVESTIGATION OF

PROACTIVE SAFETY BEHAVIOURS WITHIN

THE CONTEXT OF WORK DRIVING

Klaire Somoray

Bachelor of Behavioural Science (Psychology Honours I)

Submitted in fulfilment of the requirements for the degree of

Doctor of Philosophy

Centre for Accident Research and Road Safety - Queensland

Faculty of Health, School of Psychology

Queensland University of Technology

2019

Beyond Compliance: An Exploratory investigation of proactive safety behaviours within the context of work driving i

Keywords

Driving for work, educate, feedback inquiry, fixing safety issues, fleet safety,

intervene, leadership, meta-analysis, motivation, occupational health and safety,

organisational psychology, proactive safety behaviour, safety citizenship, safety

climate, safety compliance, safety participation, safety proactivity, structural equation

modelling, vehicle maintenance, voice, work drivers, work driving context, work

proactivity.

ii Beyond Compliance: An Exploratory investigation of proactive safety behaviours within the context of work driving

Abstract

Road trauma is the main cause of work-related death in Australia. From 2002 to

2013, two-thirds of work-related deaths in the Australian workforce were attributed to

vehicle incidents on public roads and worksites. Similar trends were also found

overseas, suggesting that work-related road trauma is a significant public health issue

that needs to be addressed. However, organisations often fail to manage the risks

associated with work driving. Furthermore, current research on work driving safety

often focus on employees’ risky driving behaviours, placing a strong emphasis on

crashes and traffic violations as measures of work drivers’ safety performance. Yet,

researchers within the general occupational health and safety field (OHS) are already

focusing on proactive approaches to work safety management. For instance, the

concept of Proactive safety behaviours is currently being used as a leading safety

performance indicator within the OHS area of research.

The current program of research provides an exploratory investigation of

proactive safety behaviours within the context of work driving (PSB-WD). PSB-WD

is defined as: 1) behaviours that improve the context of the work environment to be

more supportive of safety – these behaviours that may not directly contribute to

workplace safety, but facilitate an environment that supports safety (i.e., behaviours

that create a positive environment for safety); 2) behaviours that aim to improve

workplace safety that cannot be forced (i.e., self-starting); and 3) behaviours that are

change-oriented which aim to improve the current workplace safety practices.

Using the available research on proactive safety behaviours and related

constructs, this research program developed a measurement and model of PSB-WD.

The current program of research comprises one meta-analysis and three quantitative

studies, which are briefly outlined below:

Study 1: Meta-analytic review of the current literature that identified the

contextual and cognitive mechanisms as antecedents of proactive safety

behaviours and related constructs. Data was extracted from 71 studies. This

review helped informed the research model that was proposed and utilised

for the current program of research.

Beyond Compliance: An Exploratory investigation of proactive safety behaviours within the context of work driving iii

Study 2: A self-report measure for PSB-WD was developed using a

deductive approach to scale development. First, an item pool was generated

and later assessed by an expert panel (n = 5). This process forms the first part

of Study Two (Study 2a). Then the scale was pilot tested using N = 43

employees who drive for work (Study 2b). This study initially assessed the

reliability and dimensionality of the scale. Principal component analysis

(PCA) was used for this initial assessment. Then, the dimensionality of the

PSB-WD scale was further tested using a survey (via online and hardcopy)

with a sample of N = 300 employees who drive for work (Study 2c).

Exploratory factor analysis (EFA) and confirmatory factor analysis (CFA)

were utilised for this further assessment.

Study 3: Using the same sample from Study 2c, the proposed PSB-WD

model was examined using structural equation modelling (SEM).

For Study 1, existing studies on proactive safety behaviours and related

constructs were reviewed. The review specifically focused on the contextual and

proximal antecedents of proactive safety behaviours using the safety proactivity model

by Curcuruto and Griffin (2017). The calculations for the meta-analysis were

conducted using the metafor package in R. More specifically, the H&S-type method

within the metafor package, which is based on Hunter and Schmidt’s (2014) meta-

analysis approach, was specifically used to calculate the pooled effect size estimates.

The meta-analytic calculations were conducted on 71 studies. Within the

identified contextual antecedents, safety climate, leadership, perceived organisational

support, trust, work autonomy showed significant effects on proactive safety

behaviours. Within the identified proximal antecedents, safety knowledge, safety

motivation, self-efficacy, perceived control and promotion focus also showed

significant effects on proactive safety behaviours. Out of the 71 studies reviewed, one

study examined workers’ participation in road safety practices, and another study

investigated safety voice using a sample of work drivers.

For Study 2, a deductive approach to scale development was conducted. The

scale development involved 3 phases: 1) item pool generation, 2) expert panel, and 3)

pilot testing of the scale. First, an item pool for the PSB-WD scale was generated using

the current measures on proactive safety behaviours and related constructs. Existing

iv Beyond Compliance: An Exploratory investigation of proactive safety behaviours within the context of work driving

measures on general work proactivity were also consulted. These items were then

adapted to suit the work safety driving context. Using a nominal group technique, the

research candidate along with her supervisory team ranked and categorised the items

into possible dimensions of proactive safety behaviours. Second, an expert panel (N =

5) was approached to give further assessment on the content validity and clarity of the

items. These two-step process (item generation and expert panel) forms the first part

of Study 2 (Study 2a). The second phase of Study 2 involved the pilot testing of the

scale using a sample of employees who drive for work (N = 43). PCA was utilised for

the initial assessment of the scale and found eight potential dimensions of PSB-WD

scale, which included: feedback inquiry, helping / volunteerism, fixing responsibility,

vehicle maintenance, protection, intervene, knowledge sharing and voice.

For Study 2c, the dimensions of the PSB-WD scale were further assessed using

EFA and CFA using a sample of N = 300 employees who drive for work. The sample

completed the survey either online or via a hardcopy. The psychometric properties of

the scale were also examined. The results of the EFA demonstrated six dimensions of

PSB-WD and were confirmed using CFA. The final dimensions comprised feedback

inquiry, fixing responsibility, vehicle maintenance, intervene, educate, and voice. The

PSB-WD scale also showed good internal reliabilities, convergent validity and

divergent validity.

For Study 3, SEM was conducted to assess the proposed research model using

the same sample from Study 2c. It was hypothesised that safety climate

(organisational-level and group-level) and leader-member exchange (LMX) would

have significant relationships with employees’ proactive safety behaviours when

driving for work. Safety climate and LMX formed the contextual antecedents of the

research model. Furthermore, it was also hypothesised that anticipation focus,

perceived control and felt responsibility would also have significant relationships with

employees’ proactive safety behaviours when driving for work. These variables

formed the proximal antecedents of the model. To test the divergent validity of the

model, safety compliance and its associated proximal antecedent, preventative focus,

was added to the proposed research model.

The findings revealed that group-level safety climate, LMX, anticipation focus,

perceived control and felt responsibility demonstrated significant relationships with

work drivers’ proactive safety behaviours. More specifically, the results showed that

Beyond Compliance: An Exploratory investigation of proactive safety behaviours within the context of work driving v

group-level safety climate has a significant indirect effect on work drivers’ proactive

safety behaviours via anticipation focus, perceived control and felt responsibility,

while LMX showed a significant direct effect on work drivers’ proactive safety

behaviours. Anticipation focus, perceived control and felt responsibility also showed

significant direct effects on work drivers’ proactive safety behaviours.

On the other hand, organisational-level safety climate did not demonstrate any

significant effects on proactive safety behaviours, but it showed a significant direct

effect on safety compliance behaviours. Preventative focus also showed a direct effect

on safety compliance behaviours, but not on proactive safety behaviours.

The current program of research appealed for a proactive measure of safety

performance within the work driving context, taking insights from research on

proactive safety behaviours from the fields of traffic safety, general organisational

psychology and OHS. The PSB-WD measure that was developed could be used as a

complementary measure of behaviour-based safety performance within the context of

work driving. Furthermore, this research program also provided a model on how

organisations can engage their work drivers and management to be more proactive in

managing risks while driving for work. More specifically, the current program of

research showed that group-level safety climate, the quality of leader’s relationship

with employees, as well as employee’s propensity to anticipate risk, perceived control

and felt responsibility towards work driving safety is critical to work drivers’

engagement with proactive safety behaviours. The program of research has made

significant theoretical and practical contributions by exploring the possible application

of the proactive safety behaviour construct within the context of work driving.

vi Beyond Compliance: An Exploratory investigation of proactive safety behaviours within the context of work driving

Table of Contents

Keywords .............................................................................................................................. i Abstract ................................................................................................................................ ii Table of Contents ................................................................................................................ vi List of Figures ..................................................................................................................... ix List of Tables ....................................................................................................................... x List of Abbreviations .......................................................................................................... xi List of Publications Resulting from the PhD Research Program ....................................... xii List of Publications Outside of the PhD Program ............................................................. xiii Statement of Original Authorship ...................................................................................... xv Acknowledgements ........................................................................................................... xvi

Chapter 1: Introduction .................................................................................................... 1 1.1 Introductory Statement ............................................................................................ 1 1.2 Study Background and Rationale ............................................................................ 1

1.2.1 The Issue of Work Driving ................................................................................. 1 1.2.2 Current Work Driving Safety Management ....................................................... 4

1.3 A Need for an Alternative Framework ................................................................... 7 1.3.1 Older Models of Safety Management AKA Safety-I ......................................... 9 1.3.2 Newer Models of Safety Management AKA Safety-II ...................................... 9

1.4 A Proactive Measure of Safety Performance Within the Work Driving Context . 12 1.5 Chapter Summary and Key Learnings .................................................................. 15

Chapter 2: Proactive Safety Behaviours – A Review on Origins and Definitions ......... 16 2.1 Introductory Statement .......................................................................................... 16 2.2 Background and Origins of Proactive Safety Behaviour ...................................... 16

2.2.1 Contextual Performance ................................................................................... 18 2.2.2 Organisational Citizenship Behaviour .............................................................. 19 2.2.3 Proactivity ......................................................................................................... 21 2.2.4 Bringing it all together ...................................................................................... 23

2.3 Similar Concepts in Traffic Safety Research ........................................................ 24 2.4 Chapter Summary and Key Learnings .................................................................. 27

Chapter 3: Overview of Research Program ................................................................... 29 3.1 Introductory Statement .......................................................................................... 29 3.2 Research Questions, Aim and Objectives ............................................................. 30 3.3 Demarcation of Scope ........................................................................................... 31 3.4 Thesis Outline ....................................................................................................... 32

Chapter 4: Meta-analytic Review of Contextual and Proximal Antecedents to Proactive Safety Behaviours .............................................................................................................. 34 4.1 Introductory Statement .......................................................................................... 34

4.1.1 Purpose of the Study ......................................................................................... 34 4.1.2 Scope of the Study ............................................................................................ 35

4.2 Framework for Modelling Proactive Safety Behaviours ...................................... 36 4.2.1 Contextual Antecedents .................................................................................... 37

Safety Climate ........................................................................................................... 37 Leadership, Perceived Organisational Support and Organisational Trust ................. 40 Work Demands and Work Autonomy ....................................................................... 45

4.2.2 Proximal Antecedents ....................................................................................... 47 Safety Knowledge and Safety Motivation ................................................................. 48 Regulatory Focus, Self-efficacy and Perceived Control ........................................... 49

Beyond Compliance: An Exploratory investigation of proactive safety behaviours within the context of work driving vii

Relevance to work driving safety ............................................................................. 50 4.2.3 Summary of the Literature ............................................................................... 51

4.3 Method .................................................................................................................. 51 4.3.1 Search Strategy ................................................................................................. 51 4.3.2 Inclusion and Exclusion Criteria ...................................................................... 53 4.3.3 Coding .............................................................................................................. 55 4.3.4 Meta-analysis calculations ............................................................................... 55

4.4 Results .................................................................................................................. 57 4.4.1 Study Characteristics ........................................................................................ 57 4.4.2 Meta-analytic Correlations ............................................................................... 58 4.4.3 Contextual Antecedents ................................................................................... 66 4.4.4 Proximal Antecedent ........................................................................................ 67 4.4.5 Selected Studies Relevant to Work Driving Safety .......................................... 67

4.5 Discussion ............................................................................................................. 68 4.5.1 Practical Implications ....................................................................................... 71 4.5.2 Limitations and Future Directions .................................................................... 72 4.5.3 Chapter Summary and Key Learnings ............................................................. 72

Chapter 5: Development and Testing of Proactive Safety Behaviour Measure within the Work Driving Context ....................................................................................................... 74 5.1 Introductory Statement ......................................................................................... 74 5.2 Purpose of the Study ............................................................................................. 76 5.3 Method and Results of Study 2a – Item Generation and Expert Panel ................. 76

Item Pool Generation ................................................................................................ 77 Expert Panel .............................................................................................................. 79

5.4 Method and results of Study 2b – Pilot Study ...................................................... 81 Method ...................................................................................................................... 81 Results ....................................................................................................................... 81

5.5 Summary of Study 2a and 2b ............................................................................... 86 5.6 Method and Results of Study 2c ........................................................................... 88

5.6.1 Method ............................................................................................................. 88 5.6.2 Results of Study 2c ........................................................................................... 88

Overview of Data Analysis ....................................................................................... 88 Assessment of Missing Data ..................................................................................... 89 Participant Demographics and Work Driving Exposure ........................................... 89

5.6.3 Exploratory Factor Analysis ............................................................................. 91 Assumption Checking ............................................................................................... 91 Initial Exploratory Faction Analysis ......................................................................... 91 Final Exploratory Factor Analysis ............................................................................ 94

5.6.4 Confirmatory Factor Analysis .......................................................................... 96 Overview of Confirmatory Factor Analysis .............................................................. 96 Results of the Confirmatory Factor Analysis ............................................................ 97 Validity Assessment.................................................................................................. 99

5.7 Overall Discussion of Study 2 ............................................................................ 100 5.7.1 Theoretical and Practical Implications ........................................................... 103 5.7.2 Limitations and Future Studies ...................................................................... 104 5.7.3 Concluding Remarks ...................................................................................... 105

Chapter 6: Model Testing of Proactive Safety Behaviours within the Work Driving Context (Study 3)............................................................................................................. 106 6.1 Introductory Statement ....................................................................................... 106

6.1.1 Contextual Antecedents ................................................................................. 107 Safety Climate ......................................................................................................... 107 Leader-Member Exchange. ..................................................................................... 109

6.1.2 Proximal Antecedents .................................................................................... 111 Proactive Future Orientation – Anticipation-Focus ................................................ 111

viii Beyond Compliance: An Exploratory investigation of proactive safety behaviours within the context of work driving

Proactive Motivational States – Can Do and Reason To ........................................ 112 6.2 Non-Proactive Antecedent, Mediators and Safety Compliance .......................... 114 6.3 Study Hypotheses ................................................................................................ 117 6.4 Method ................................................................................................................ 118 6.5 Measures ............................................................................................................. 118

6.5.1 Proactive Safety Behaviours within the Work Driving Context (PSB-WD). . 118 6.5.2 Safety Climate Survey – Short Form. ............................................................. 119 6.5.3 Leader-Member Exchange (LMX). ................................................................ 119 6.5.4 Anticipation Focus, Perceived Control and Felt Responsibility. .................... 119 6.5.5 Prevention Focus and Safety Compliance. ..................................................... 120

6.6 Data Analysis ...................................................................................................... 121 6.6.1 Overview of Data Analysis ............................................................................. 121

6.7 Results ................................................................................................................. 121 6.7.1 Treatment of Missing Data ............................................................................. 121 6.7.2 Aggregation of Items ...................................................................................... 122 6.7.3 Assumption Checking ..................................................................................... 122

Common Method Variance Test ............................................................................. 123 6.7.4 Description of Data ......................................................................................... 124 6.7.5 Who are more likely to engage in Proactive Safety Behaviours? ................... 126 6.7.6 Hypothesis Testing using Structural Equation Modelling .............................. 128

Test of Direct and Indirect Effects .......................................................................... 128 Hypothesis on the Non-Proactive States, Motivation and Outcomes ...................... 130

6.8 Discussion ........................................................................................................... 131 6.8.1 Limitations and Future Research .................................................................... 134 6.8.2 Chapter Summary and Key Learnings ............................................................ 135

Chapter 7: Discussion and Conclusions ....................................................................... 137 7.1 Introductory Statement ........................................................................................ 137

7.1.1 Overall Findings in Relation to RQ1 .............................................................. 138 7.1.2 Overall Findings in Relation to RQ2 .............................................................. 142 7.1.3 Overall Findings in Relation to RQ3 .............................................................. 143 7.1.4 Theoretical Implications and Contributions ................................................... 144 7.1.5 Strengths and Practical Implications .............................................................. 147 7.1.6 Limitations and Suggestions for Future Research .......................................... 150 7.1.7 Concluding Remarks ...................................................................................... 152

Bibliography .................................................................................................................... 153

Appendices ....................................................................................................................... 175

Beyond Compliance: An Exploratory investigation of proactive safety behaviours within the context of work driving ix

List of Figures

FIGURE 2.1 OVERVIEW OF THE HISTORICAL ORIGINS OF PROACTIVE SAFETY RESEARCH ....................... 18 FIGURE 2.2 FIELDS OF RESEARCH THAT HELPED INFORM THE PROACTIVE SAFETY BEHAVIOUR

CONSTRUCT WITHIN THE WORK DRIVING CONTEXT AND THE FOUNDATION OF THE CURRENT

PROGRAM OF RESEARCH .............................................................................................................. 27 FIGURE 3.1 OUTLINE OF THE PROGRAM OF RESEARCH AND ITS PRESENTATION IN THIS THESIS ............. 33 FIGURE 4.1 CURCURUTO AND GRIFFIN’S (2018) MODEL OF ANTECEDENTS OF SAFETY PROACTIVITY. THE

GRAY SHADED BOXES WERE THE FOCUS OF THE CURRENT META-ANALYSIS. ............................... 36 FIGURE 4.2 SEARCH STRATEGY FOR THE SYSTEMATIC REVIEW AND META-ANALYSIS .......................... 53 FIGURE 5.1 HYPOTHESISED DIMENSIONS OF PROACTIVE SAFETY BEHAVIOURS WITHIN THE WORK

DRIVING CONTEXT (PSB-WD; LEFT) AND THE DIMENSIONS EXTRACTED FROM PCA USING THE

PILOT STUDY DATA (RIGHT). ........................................................................................................ 87 FIGURE 5.2 RE-SPECIFIED MEASUREMENT MODEL OF PSB-WD SCALE ................................................. 99 FIGURE 5.3 HYPOTHESISED DIMENSIONS OF PROACTIVE SAFETY BEHAVIOURS WITHIN THE WORK

DRIVING CONTEXT FROM THE ITEM GENERATION (PSB-WD; LEFT) AND THE DIMENSIONS

EXTRACTED FROM PCA USING THE PILOT STUDY DATA (MIDDLE) AND DIMENSIONS EXTRACTED

FROM THE EFA AND CFA (LEFT)............................................................................................... 102 FIGURE 6.1 RESEARCH MODEL ............................................................................................................ 106 FIGURE 6.2 RESEARCH MODEL WITH HYPOTHESES .............................................................................. 117 FIGURE 6.3 RESPECIFIED MODEL AFTER CONSULTING THE MODIFICATION INDICES ............................ 129 FIGURE 6.4 RESPECIFIED MODEL WITH STANDARDISED REGRESSION ESTIMATES, ONLY SHOWING THE

SIGNIFICANT PATHS. .................................................................................................................. 130 FIGURE 6.5 RESEARCH MODEL WITH PREVENTION FOCUS AND SAFETY COMPLIANCE TO TEST THE

DISCRIMINANT VALIDITY ........................................................................................................... 131

x Beyond Compliance: An Exploratory investigation of proactive safety behaviours within the context of work driving

List of Tables

TABLE 1.1 TRADITIONAL MODELS OF SAFETY VERSUS HRO AND RESILIENCE ENGINEERING. THIS

TABLE WAS ADAPTED FROM HOLLNAGEL ET AL. (2013, P. 26). .................................................... 11 TABLE 4.1 SEARCH TERMS USED FOR EACH DATABASE ......................................................................... 52 TABLE 4.2 TERMINOLOGIES AND MEASURES USED FOR THE CURRENT STUDY ....................................... 54 TABLE 4.3 STUDIES INCLUDED IN THE DATA-ANALYSIS ......................................................................... 59 TABLE 4.4 RESULTS OF THE META-ANALYSIS ........................................................................................ 65 TABLE 5.1 CHAPTER SECTION, PHASES, METHODOLOGY AND PURPOSES OF STUDY 2A, 2B AND 2C ....... 76 TABLE 5.2 FINAL PCA FOR THE PILOT STUDY SURVEY PSB-WD ......................................................... 84 TABLE 5.3 REVISED AND ADDED ITEMS ................................................................................................. 86 TABLE 5.4 DEMOGRAPHIC AND WORK INFORMATION OF PARTICIPANTS ................................................ 90 TABLE 5.5 INITIAL PAF WITH PROMAX ROTATION ................................................................................ 93 TABLE 5.6 FINAL PAF WITH PROMAX ROTATION .................................................................................. 95 TABLE 5.7 ASSESSMENT OF MODEL FIT ................................................................................................. 97 TABLE 5.8 DISCRIMINANT AND CONVERGENT VALIDITY ..................................................................... 100 TABLE 6.1 MEANS, STANDARD DEVIATIONS, CRONBACH ALPHA AND INTER-CORRELATIONS BETWEEN

THE VARIABLES .......................................................................................................................... 125 TABLE 6.2 BIVARIATE CORRELATIONS BETWEEN PROACTIVE SAFETY BEHAVIOUR VARIABLES AND

WORK DRIVING EXPOSURE AND RISK .......................................................................................... 127 TABLE 6.3 STANDARDISED REGRESSION WEIGHTS AND 95% BIAS-CORRECTED CONFIDENCE INTERVALS

WITH 10,000 BOOTSTRAP SAMPLING .......................................................................................... 129 TABLE 7.1 OVERVIEW OF DESIGN AND METHODOLOGY STRUCTURED BY THE RESEARCH QUESTIONS .. 137

Beyond Compliance: An Exploratory investigation of proactive safety behaviours within the context of work driving xi

List of Abbreviations

Abbreviation / Symbol Definition

α Cronbach’s alpha

β Beta

χ2 Chi-square

BCa Bias-corrected and accelerated bootstrap

CI Confidence Intervals

CFA Confirmatory Factor Analysis

CMB Common Method Bias

EFA Exploratory Factor Analysis

HRO High Risk Organisation

LMX Leader-Member Exchange

MVA Missing Value Analysis

OHS Occupational Health and Safety

PCA Principal Component Analysis

PAF Principal Axis Factoring

PSB-WD Proactive Safety Behaviours within the Work Driving

Context

SEM Structural Equation Modelling

xii Beyond Compliance: An Exploratory investigation of proactive safety behaviours within the context of work driving

List of Publications Resulting from the PhD Research Program

Conference Paper:

Somoray K., Newton C., Lewis I., Wishart D. (2019). Development of proactive safety behaviour scale within the work driving context. In P Arezes (Eds.), Advances in Safety Management and Human Factors (pp. 470-479) Switzerland: Springer. doi: https://doi.org/10.1007/978-3-319-94589-7_46

Beyond Compliance: An Exploratory investigation of proactive safety behaviours within the context of work driving xiii

List of Publications Outside of the PhD Program

Academic Journals:

Street, T. D., Somoray, K., Richards, G. C., & Lacey, S. J. (2019). Continuity of care for patients with chronic conditions from rural or remote Australia: A systematic review. Australian Journal of Rural Health. doi: 10.1111/ajr.12511 Street, T., Lacey, S., & Somoray, K. (2019). Employee stress, reduced productivity, and interest in a workplace health program: a case study from the Australian mining industry. International journal of environmental research and public health, 16(1), 94. doi: /10.3390/ijerph16010094

Debnath, A.K., Haworth, N., Schramm, A., Heesch, K.C., Somoray, K. (2018). Factors influencing noncompliance with bicycle passing distance laws. Accident Analysis and Prevention, 115, 137-142. doi: 10.1016/j.aap.2018.03.016c

Wishart, D., Somoray, K., & Evenhuis, A. (2017). Thrill and adventure seeking in risky driving at work: The moderating role of safety climate. Journal of Safety Research, 63, 83-89. doi: 10.1016/j.jsr.2017.08.007

Wishart, D., Somoray, K., & Rowland, B. (2017). Role of thrill and adventure seeking in risky work-related driving behaviours. Personality and Individual Differences, 104, 362-367. doi:10.1016/j.paid.2016.08.033

Somoray, K., Shakespeare-Finch, J., & Armstrong, D. (2017). The impact of personality and workplace belongingness on mental health workers’ professional quality of life. Australian Psychologist, 52, 52-60. doi: 10.1111/ap.12182

Book Chapters:

Wishart, D., Rowland, B., & Somoray, K. (2019). Safety citizenship behaviour: A complimentary paradigm to improving safety culture within the organisational driving setting. In N. J. Ward, B. Watson & T. Özkan (Eds.), Traffic Safety Culture: Theory, Measurement and Application, Howard House, Bingley: Emerald Publishing Ltd.

Conference Papers:

Licina, P., Johnston, E., Somoray, K., Rakotonirainy, A., & King, M. (2017). Fitness to drive after lumbar discectomy – A pilot study using an advanced driving simulator. In 2017 Spine Society of Australia 28th Annual Scientific Meeting, 7-9 April, 2017, Hobart, Tasmania.

xiv Beyond Compliance: An Exploratory investigation of proactive safety behaviours within the context of work driving

Somoray, K. & Wishart, D. (2016). Review of research methodologies in investigating work-related driving behaviour. In 2016 Australasian Road Safety Conference, 6-8 September 2016, Canberra, A.C.T.

Wishart, D., Somoray, K., & Rowland, B. (2016). Reducing reversing vehicle incidents in Australian fleet settings – A case study. In Stanton, Neville A., Landry, Steve, Di Bucchaianico, Giuseppe, & Vallicelli, Andrea (Eds.) Advances in Human Aspects of Transportation. Springer, 733-744.

Wishart, D., Somoray, K., & Rowland, B. (2016). A longitudinal study evaluating work driving safety interventions implemented by a number of organisations. In 2016 Australasian Road Safety Conference, 6-8 September 2016, National Convention Centre, Canberra, A.C.T.

Rowland, B., Wishart, D., & Somoray, K. (2015). Organisational driving safety systems analysis: fleet safety situational issues and system gaps. In 2015 Australasian Road Safety Conference, 14-16 October 2015, Gold Coast, QLD.

Beyond Compliance: An Exploratory investigation of proactive safety behaviours within the context of work driving xv

Statement of Original Authorship

The work contained in this thesis has not been previously submitted to meet

requirements for an award at this or 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.

Signature: QUT Verified Signature

Date: 14 June 2019

xvi Beyond Compliance: An Exploratory investigation of proactive safety behaviours within the context of work driving

Acknowledgements

The past three years have flown so fast. My journey in work driving safety

research started four years ago, when Darren asked me to work for him as a Research

Assistant at CARRS-Q. Since then, I have been fortunate to be part of various research

projects, within and outside the area of work driving safety. My passion for research,

and especially for statistics and data analysis, has flourished over the years.

This passion for research continues to grow. For a large part, I have my

supervisors, Darren, Cameron and Ioni, to thank for that. I always look forward to our

fun meetings and email chats. Cam and Ioni, you were not in the original supervisory

team, but you took me on, and provided support as if you were there from the start.

Thank you Darren for always believing in me, thank you Cam for the statistical advice

(and crude jokes), and thank you Ioni for the swift feedback, constructive comments

and constant encouragement. I could not have asked for better supervisors. I also

would like to acknowledge QUT scholarship for the financial support.

To my wonderful friends. Thank you for the continuous support and encouraging

messages, even if I've been hiding in my cave or forget to reply. Special thanks to

John, Kai, Jaggy, Evie and Ada for helping me de-stress. John, you've been there from

the beginning. You've always listened, and continue to listen, to my nerdy discussions.

To the friends I made during my PhD journey - especially Sep, Mahrokh, Wahi

and Raniya. You have been my constant rock. The laughter, the late night chats, the

pep talks, the frequent "why are we doing this again?" discussions, the dinners, the

picnics, the way we celebrate each other's small wins - I am finishing my thesis with a

big smile because of you guys. Thank you.

To Michelle. You've been my source of comfort and stability, especially in the

past few months. You always listen to my stress rants – cheering me on and brightening

up my day. Thank you for the continuous love, patience, encouragement and support.

I am so grateful to have you in my life and I look forward to our future travels.

And lastly to my family, especially to my Mom and Dad. You always

encouraged me in everything that I do. You always celebrate my achievements,

however small. Your love and guidance have kept me going and your hard work has

Beyond Compliance: An Exploratory investigation of proactive safety behaviours within the context of work driving xvii

always been my source of inspiration. I truly appreciate everything you do for me, but

I know that I can never thank you enough. Ma and Pa, I dedicate my work and

achievements to both of you.

Chapter 1: Introduction 1

Chapter 1: Introduction

1.1 INTRODUCTORY STATEMENT

The current program of research aims to explore the concept of proactive safety

behaviour as a possible indicator of safety performance within the work driving

context. This research is important because work-related road trauma is a significant

public health issue. Organisations also have the legal responsibility to ensure the safety

of their employees while driving for work. Using theoretical frameworks and

constructs from organisational psychology, occupational health and safety (OHS) and

traffic safety, the overarching aim of this program of research was to examine the

concept of proactive safety behaviours and its possible application within the work

driving context.

This chapter begins by providing the background and rationale for the research

program. Relevant literature on work driving safety is critically reviewed and

summarised, with a specific focus on the significant impact of work-related road

trauma on the public’s health and safety. The gaps within the management of work

driving safety, from a practitioner and research point of view, are also highlighted.

A need for an alternative framework for managing work driving safety is

essential. High Reliability Organisation (HRO) Theory and Resilience Engineering,

also known as Safety-II models, are specifically used as the theoretical lens to guide

the progression of the program of research. Before the discourse on these models, the

traditional models of occupational safety (Safety-I) and the shift towards the more

recent models are discussed. Then, a need for a proactive measure of safety behaviour

within the work driving context is proposed.

1.2 STUDY BACKGROUND AND RATIONALE

1.2.1 The Issue of Work Driving

Deaths and injuries resulting from a vehicle crash is a significant public health

issue that needs to be addressed (Robb, Sultana, Ameratunga, & Jackson, 2008).

Globally, road trauma is associated with 1.3 million deaths and up to 50 million non-

fatal injuries each year, costing governments an estimated 3% of their gross domestic

2 Chapter 1: Introduction

product annually (World Health Organization, 2015). In Australia, road crashes result

in an average of 1,400 deaths and 32,500 serious injuries annually with an estimated

cost of $27 billion to the economy (Department of Infrastructure Regional

Development and Cities, 2018).

These human and societal costs are tragic and unnecessary, given that the

majority of road traffic incidents are preventable (National Safety Council, 2014;

World Health Organization, 2015). However, despite the overwhelming number of

deaths and injuries resulting from a road crash, there seems to be a general sense of

complacency and acceptance towards road traffic risks (National Safety Council,

2014). It is possible that due to the repeated exposure to driving, a large majority of

people see driving as an everyday task and the associated risks of injury and fatality

are often overlooked (Banks, 2008; National Safety Council, 2014).

This complacency towards driving risks is also observed within organisations

(Banks, 2008). For instance, most organisations in high risk industries (e.g., mining

and constructions) often have zero harm policies for central work processes but often

do not have any comprehensive safety policies for driving during working hours

(Newnam, Wishart, & Newton, 2014). This lack of attention on workers’ safety while

driving is problematic, given that work-related driving incidents present the majority

of casualties and financial costs to society (Australian Transport Council, 2011).

According to the Australian Transport Council (2011), fifteen percent of the

national road toll in Australia are a result of work-related crashes. This proportion is

even higher if commuting to and from work is considered as work-related travel. On a

more concerning note, while the total number of work fatalities and serious injuries is

decreasing, the proportion of vehicle-related incidents at work has remained relatively

stable between 2003 and 2016 (Safe Work Australia, 2018b). A recent report by Safe

Work Australia (2018b) demonstrated that the proportion of vehicle crashes as a

mechanism of fatality and serious injury at work has been between 35% and 43% for

13 years1. These figures suggest that, each year, over one-third of worker fatalities are

attributed to the result of a road crash. In addition, analysis of worker fatalities from

2007 to 2016 revealed that nearly two-thirds (64%) of deaths in the Australian

1 This proportion remained relatively stable between 2003 and 2016, aside from year 2015 when the proportion was at its lowest at 28%.


workforce involved a vehicle (Safe Work Australia, 2018a). Of these vehicle-related

deaths, over 45% occurred due to a vehicle collision on public roads (Safe Work

Australia, 2018a). It is important to note that, while these data examined crashes from

both light and heavy vehicle fleets, it is likely that the majority of the crash data is

derived from heavy vehicle industries as these organisations are governed under

federal laws and regulations. Therefore, the available data may be underestimating the

extent of the issue as crashes from light vehicle fleets are not always recorded

(Stuckey, Pratt & Murray, 2013).

Other industrialised countries also face similar issues. Data analysis conducted

by the European Agency for Safety and Health at Work (2010) revealed that road

trauma accounts for 29% of worker fatalities throughout Europe. In the United

Kingdom, it was reported that 7% of the national traffic fatalities were work-related

and this number increases to 23% when commercial vehicles were involved (Murray,

2007). Road crash is also the leading cause of work-related deaths in the United States

(Pratt, 2003). In 2014, nearly a quarter of fatal work-related injuries suffered by

American employees occurred on the road (Bureau of Labor Statistics, 2014). From

these reports, it is clear that driving is one of the most hazardous job activities for

workers to undertake (Haworth, Greig, & Wishart, 2008; Wishart, 2015).

In addition to the tragic loss of lives, Wheatley (1997) reported that occupational

driving incidents also cost the Australian government approximately 1 billion to $1.5

billion annually, while a study by Davey and Banks (2005) estimated an average of

$28,000 per reported work-related vehicle incident in total insurance costs for

companies. Motor vehicle incidents are also over-represented in Australian workers’

compensation claims. From 2010 to 2011, vehicle incidents were the most common

cause of Australian compensated fatalities, accounting for almost one-third of all

compensated deaths (Safe Work Australia, 2013).

Clearly, there are notable benefits to conducting research within the work-related

road safety domain. Many lives and financial costs could be saved if risks related to

work driving were properly managed and adequate measures were put in place

(Wishart, 2015). In order to achieve high levels of safety while driving for work,

management and workers must be proactive in addressing the risks related to work

driving. Research should focus on investigating the leading indicators of safety

performance that aim to improve safety in the work driving domain instead of reacting


to past events (Newnam et al., 2014). While research over the past decade is starting

to focus on leadership practices and system-based approaches to understanding work

driving safety (e.g., Newnam & Goode, 2015; Newnam & Oxley, 2016), the majority

of the research and practice in this field still tend to focus on employees’ involvement

in crashes and traffic violations as measures of safety performance when driving for

work (Newnam et al., 2014). Furthermore, these incidents are often used by

organisations to inform their risk management process which, in turn, provides a

superficial antidote to a problem without fully understanding the cause (Wishart,

2015). For example, a crash is attributed to speeding, but the contributing factors (such

as speeding to meet clients on time) may not often be investigated. Arguably, there is

a need to shift from the reactive perspective that currently dominates the field of work

driving safety to that of a more proactive risk management approach.

While research and practice in the work driving safety field are still primarily

focused on lagging indicators of safety (Newnam et al., 2014), the general OHS field

is already placing greater emphasis on concepts that go beyond accidents, injuries and

minimum safety compliance to further improve safety at work (Curcuruto & Griffin,

2017). In order to adequately address work driving safety, there is a clear need for a

perspective shift in the way that researchers and organisations currently examine risk

management when driving for work (Wishart, 2015).

1.2.2 Current Work Driving Safety Management

Workplace health and safety legislation in Australia also requires organisations

to be responsible for the safety of their employees, and this includes driving for work

(Robb et al., 2008; Safe Work Australia, 2016a). Within this legislation, vehicles used

for work are defined as part of a workplace and driving for work is considered a work-

related task (Austroads, 2018). Recently, the industrial manslaughter provisions in the

Work Health and Safety Act (2011) makes it a criminal offence for employers and

senior officers in Queensland to negligently cause the death of a worker while

undertaking work-related tasks or during working hours (including a work break;

Queensland Government, 2018). Since work-related travel is considered a job task,

employers can also be held liable for not implementing their legal obligations and

ensuring a safe system of work when employees use motor vehicles for work (Wishart,

2015).


Furthermore, a large proportion of road commute within Australia is work-

related. In 2012, 20% of the registered vehicles in Australia include commercial

vehicles, trucks and buses and 33.5% of the annual mileage travelled by Australians

are for business purposes (Australian Bureau of Statistics, 2012). However,

organisations often fail to mitigate the risks associated with work driving with the same

level of diligence as other work-related hazards (Wishart, 2015).

There is a significant contrast in the management of risks between general work

tasks and work driving. For example, risk management within mining or construction

industries operate under the ‘zero harm’ framework (Wishart, 2015). Within this

framework, any activity that compromises safety is not tolerated and unacceptable.

Furthermore, these industries allocate appropriate resources in building workplace

health and safety management teams, training and awareness programs, and safety

consultations to ensure compliance with safety procedures and policies (Wishart,

2015). However, when managing employees’ safety when driving for work,

organisations often lack policies and procedures on this particular work activity

(Haworth et al., 2008; Murray, 2007; Murray, Newnam, Watson, Davey, & Schonfeld,

2003; Wishart, 2015). Wishart (2015) further argued that risk management on work

driving often is inadequately resourced (e.g., lack of personnel and education

programs) and often attract lower levels of diligence. Indeed, it is not surprising that

safe driving practices are usually compromised and employees who drive for work

suffer from more severe and frequent crashes at work (Stuckey, Lamontagne, Glass,

& Sim, 2010).

If organisations do implement risk management for work driving, they often take

precautions following a crash or violation, and the efficacy of this approach is limited

by the analysis of incidents that happened in the past (Wishart & Davey, 2004). In a

project that evaluated work-related driving safety practices in Australia, Murray et al.

(2003, p. 10) stated that senior managers “tend to focus on fleet safety as a reactive

response to a range of negative events such as being involved in a fatality or very

expensive crash” (p. 10). Murray et al. (2003) also found that management of work-

related crashes typically centre on the employee – usually putting the blame on the

work driver when an incident occurs (Murray et al., 2003). Wishart and Davey (2004,

p. 2) also reported that organisations often take a “silver bullet approach aimed at

developing and implementing a single countermeasure or intervention strategy to


encompass and address all fleet related road safety issues” (p. 2). This reactive

approach to safety often emphasises the fault of those involved, instead of encouraging

a deeper examination of the organisational processes that may have played a major

role in the occurrence of the incident (Hudson, 2003).

Recently, work led by Newnam (e.g., Newnam & Goode, 2015; Newnam,

Goode, Salmon & Stevenson, 2017; Warmerdam, Newnam, Sheppard, Griffin &

Stevenson, 2017; Newnam & Oxley, 2016; Newnam, Lewis & Watson, 2012), Salmon

(e.g., Salmon, McClure & Stanton, 2012; Salmon and Lenne, 2015) and has been

applying a systems-based approach to road safety. This systems-based approach

challenges the reductionist approach of ‘blaming the driver’ that focuses on unsafe

driving behaviours and instead examines the road transportation system as a whole

(Newnam and Goode, 2015). For instance, Newnam and Goode (2015) stated that the

road freight system has the characteristics of a complex sociotechnical system.

Consequently, a crash caused by fatigue may indicate the driver’s disregard of fatigue

management policies and/or could indicate the supervisor’s lack of involvement on

journey management or the company’s policies on production over safety (Newnam

and Goode, 2015). While these recent studies look at the complexity of the road

transportation systems as factors of road incidents (rather than just ‘blaming the

driver’), these studies still look at crashes as a performance indicator.

Furthermore, current risk management in work driving promotes a strong focus

on compliance with standard rules and regulations rather than promoting a proactive

approach to safety (Murray et al., 2003). This approach is based on the premise that

safe driving is mostly reliant on workers’ compliance with traffic laws and

organisations’ safety policies and procedures (if available). While compliance with

safety policies reduces the risk of accidents and injuries at work (Christian, Bradley,

Wallace, & Burke, 2009), research suggests that such incidents continue to occur even

when workers comply with rules and regulations (Hu, Yeo, & Griffin, 2018). These

incidents occur because workers comply with safety procedures in different manners,

and not all standards and procedures in place are appropriate to work activities (Hu et

al., 2018). Hu et al. (2018) argued that blindly following rules instead of doing the task

safely can be as dangerous as non-compliance behaviours. Thus, it is important to

examine a proactive approach in managing risks while driving for work, not only for


the safe and successful operations of organisations, but also for the safety of the

community.

1.3 A NEED FOR AN ALTERNATIVE FRAMEWORK

While research on proactive safety management approaches within the general

OHS) field is readily available, there is very limited research on this topic within the

work driving area (Mooren, 2016). Research on work driving safety is already limited

in the first place (Wishart, 2015) and Mooren (2016) argued that there seemed to be

little crossover between the general OHS research and fleet safety. For instance, the

earliest accounts of driver management programs in the literature appeared to be

conducted in 1995 by Adams-Guppy and Guppy (1995) and Kedjidjian (1995)2. Using

a sample of British company car drivers, Adams-Guppy and Guppy (1995) found that

the participants who reported to exceed speeding limits placed more importance on

getting to appointments on time. Therefore, they recommended for organisations to

apply realistic scheduling to reduce time-urgency factors and to reduce employees’

speeding behaviours. Kedjidjian (1995), on the other hand, recommended defensive

driver techniques, driver education and installation of vehicle safety features to

manage employees’ safety while driving for work.

Aside from these studies, it was only around the turn of the century (2000’s) that

interest piqued on driver management programs that target other factors, apart from

driving error, as potential causes of work driving incidents (Haworth et al., 2008).

Previously, management of work driving were based on the ‘carrots-and-stick’

approach. These programs usually involved giving rewards for the absence of traffic

violations and applying punishments when workers get into road incidents. On the

other hand, more recent studies within the OHS field is moving away from the ‘carrots-

and-stick approach’, and human errors and violations are being seen as the

consequence of organisational shortcomings rather than the cause of safety incidents.

Findings from the OHS field are, however, not being implemented within the area of

work driving safety (Mooren, 2016).

2 These findings are based on Haworth, Tingvall and Kowadlo’s (2000) review of best practices in fleet safety.


There is a clear need for an alternative framework for managing work driving

risks. This framework should place less focus on lagging indicators3 and more focus

on identifying and recognising leading indicators (Lofquist, 2010). Lofquist (2010)

argued that this framework would require a new set of soft metrics (i.e., attitudes and

behaviours) captured and reported by work drivers which could help anticipate risks

before incidents occur. To guide the current program of research, the frameworks of

HRO and Resilience Engineering (also known as Safety-II) had been utilised.

Before starting the discussion on these frameworks, however, it is important to

note that several safety models exist within the general road safety context (e.g.,

Haddon Matrix; Safety Systems Model). Still, given the occupational nature of the

current program of research, it is arguably more appropriate to examine the theoretical

developments within the general OHS field in identifying and conceptualising leading

indicators that could be used within the work driving setting.

Several researchers have also proposed frameworks for improving work driving

safety (e.g., Stuckey et al.’s [2010] Occupational Light Vehicle Systems Model;

Murray, Watson, King, Pratt & Darby’s [2014] Haddon Matrix for work driving).

While these models are useful in identifying the different factors that could influence

work drivers’ safety from the individual (e.g., drivers’ characteristics) and broader

level (e.g., organisational, community) perspectives, these models do not provide

descriptions on what indicators could be measured to assess the work drivers’ safety

performance before a crash occurs. These models are still focused on lagging

indicators, rather than leading indicators. On the other hand, Safety-II models offer an

alternative approach to understanding safety and provide a basis on what these leading

indicators could be. Therefore, these safety frameworks were utilised as the theoretical

lens which guided the progression of the current research program. The purpose of

applying these frameworks is to explore the possible ways to measure leading

indicators of safety performance within the work driving context. Furthermore, these

frameworks would provide further rationale on the importance of the current program

of research.

3 Lagging indicators measure the safety performance of an organisation based on past incidents (e.g., frequency and severity of injuries) while leading indicators measure safety events that may happen in the future (O’Neill, Martinov-Bennie, Cheung, & Wolfe, 2013).


To fully understand the concept of Safety-II, a discussion of older safety models

or Safety-I is first required.

1.3.1 Older Models of Safety Management AKA Safety-I

Traditional models of safety are rationalist and prescriptive, often attributing the

cause of work accidents to human failure and putting little emphasis on the role of

broader organisational structures and social contexts in accident causation (e.g.,

Heinrich’s Domino Theory; Toft, Dell, Klockner, & Hutton, 2012). Within the

traditional models of safety management, the main objective is to keep accidents and

risks to a minimum level, with rules seen as “essentially static and in a sense linear”

(Hale & Borys, 2013, p. 211). Hale and Borys (2013) argued that this approach to

safety is based on major accidents where rule violation is identified as the cause. As a

response, management calls for more stringent and extensive rules to prevent

recurrence of such events. Close supervision, safety rules and regulation, employee

education and accident analysis are encouraged to limit the “unsafe acts” performed

by employees (Safe Work Australia, 1996).

These models of safety have persisted over a number of decades (from early

1930s to 1980s) and some of its influence is still evident in the way that OHS is

managed in organisations (Hale & Borys, 2013). However, these models are

problematic as they paint a simplistic approach to accident causation.

In the 1990s, Reason (1990) proposed the Swiss-cheese model of accident

causation, which is influential in shifting the view on occupational safety. Reason

(1990) argued that, in a complex system, hazards are prevented by a series of barriers.

However, each barrier has its own unintended weakness or “holes” (like a stack of

Swiss cheese). According to this model, accidents occur due to the failure of

recognising these weaknesses and failure to put up defences to stop hazards and errors

from passing through. While the Swiss-cheese model has its critics, it was influential

in changing the perspective of humans as the main cause of accidents to recognising

the role of systems and the work environment (as well as its interaction with human

behaviour) in accident causation and prevention (Toft et al., 2012).

1.3.2 Newer Models of Safety Management AKA Safety-II

Since then, the study of work environment, organisational culture and safety

management system have dominated the literature (Pillay, Borys, Else, & Tuck, 2010).


Instead of focusing on why things go wrong, safety researchers are starting to look at

how things go right. Recently, most OHS research has begun investigating how

organisations still manage to be safe and productive in the face of economic challenges

and ongoing disruptions to production. The High Reliability Organisations (HRO)

theory provides a comprehensive description on how high risk organisations maintain

safety for long periods even when operating in complex and hazardous environments

(Glendon, 2011; Lekka, 2011). Developed by researchers from the University of

California, the HRO theory maintains a positive view on the nature of accidents in

complex systems by arguing that organisations can become more reliable by creating

a ‘just safety culture’ that focuses on identifying system failures rather than blaming

the individuals. Within this perspective, errors could still occur, but can be prevented

through a combination of organisational design, culture, management and human

decision making.

Organisations that operate within the HRO framework are constantly pre-

occupied with failure so that they can anticipate areas of potential failure and cope

effectively when errors do occur (Glendon, 2011). This feature of HRO is often

described as “collective mindfulness”. Other characteristics of HRO include a strong

learning orientation, where workers are continuously trained to recognise and manage

errors and willingness to learn from mistakes (Glendon, 2011). In Weick and Roberts’s

(1993) observational study, they found that operators in aircraft carriers tackle risk and

eliminate errors by relying on their past training and comprehension of the situation.

There was also an emphasis on checks and procedures to ensure that workers could

correct errors before they escalate into crisis (Glendon, 2011).

Resilience engineering, on the other hand, was coined by Hollnagel, Woods,

and Leveson (2007) to describe a safety management approach where organisations

are able to operate under pressure while being productive and still achieving high-

levels of safety. Resilience engineering seeks to enhance the ability of organisations

by creating proactive risk models at all levels of management, promote processes that

are robust and flexible and direct resources proactively despite disruptions to

production and economic pressures (Dekker, Hollnagel, Woods, & Cook, 2008). The

HRO theory shares similar concepts and principles with resilience engineering, where

resilient organisations have the capacity to bounce back after experiencing safety

challenges and mishaps.


For an organisation to be safe, reliable and resilient, management should

discourage a blind following of rules and instead foster proactive safety behaviours

and safety awareness among their workers (Hale & Borys, 2013). Workers are

encouraged to recognise critical situations and have the ability to identify a safe action

beyond safety regulations (Dekker et al., 2008). Various OHS researchers have argued

that the framework of HRO and resilience engineering should be applied in other

transportation modes, such as vehicle fleets, when managing risks (Glendon, 2011;

Lofquist, 2017; Naveh & Marcus, 2002). For instance, Naveh and Marcus (2002)

argued that a high percentage of traffic crashes can be attributed to drivers’ error and

lack of attention. In nuclear power plants and commercial aviation, the operators’

‘errors’ are thoroughly scrutinised and analysed to determine how disasters can be

prevented. This information is then fed back to operators and system designers to

optimise the safety system networks. Similar systems should be in place within the

work driving contexts, where both the work drivers and their supervisors need to be

more aware of road hazards and what decisions to take to avoid these risks.

Therefore, these frameworks were utilised as the ‘conceptual lens’ that guided

the progression of this research4. Table 1.1 shows the differences between the

traditional models of safety compared to Safety-II models. As shown in Table 1.1, the

current program of research defined safety as “many things as possible go right” and

safety is managed by constant anticipation of possible errors and issues. Rather than a

liability, workers are also considered as a resource, who are “necessary for system

flexibility and resilience” (Hollnagel et al, 2013, p. 26). From these Safety-II concepts,

the current research program proposed proactive safety behaviours as a metric of

leading safety indicators within the work driving context.

Table 1.1 Traditional Models of Safety versus HRO and Resilience Engineering. This table was adapted from Hollnagel et al. (2013, p. 26).

Traditional Models of Safety aka

Safety-I

HRO and Resilience Engineering aka

Safety-II

Definition of Safety That as few things as possible go wrong.

That as many things as possible go right.

4 It is important to acknowledge that while the focus of the current program of research was on the behaviours of the work drivers (and Resilience Engineering and HRO tend to focus on the wider system), the program also considered the other factors within the work environment (leadership and safety climate) that could impact on proactive safety behaviours.


Safety management principles

Reactive, respond when something happens or is categorised as an unacceptable risk.

Proactive, continuously trying to anticipate developments and events.

View of human factor in safety management

Humans are predominantly seen as a liability or hazard.

Humans are seen as a resource necessary for system flexibility and resilience.

Accident investigation

Accidents are caused by failures and malfunctions. The purpose of an investigation is to identify the causes.

Things basically happen in the same way, regardless of the outcome. The purpose of an investigation is to understand how things usually go right as a basis for explaining how things occasionally go wrong.

Risk Assessment

Accidents are caused by failures and malfunctions. The purpose of an investigation is to identify causes and contributory factors.

To understand the conditions where performance variability can become difficult or impossible to monitor and control.

1.4 A PROACTIVE MEASURE OF SAFETY PERFORMANCE WITHIN THE WORK DRIVING CONTEXT

Research and practice within work driving safety, however, is lagging behind

other work hazard risk management (Wishart, Bevan, & Somoray, 2019)5. As

mentioned previously, there is currently a strong emphasis on managing work driving

after an incident has occurred or only focusing on meeting the minimum requirements

for safety. Within the work driving context, safety performance is still viewed using

traditional models, with greater emphasis on frequency and severity of crashes and

traffic violations (Wishart et al., 2019). These measures are ineffective indicators of

safety performance because accidents and injuries only reflect occurrences of failure.

In addition, there are difficulties with assessing on-road behaviours,

organisational crashes and traffic offences. For instance, research on driving behaviour

typically utilise self-report measures of crash involvement, which are susceptible to

biases (e.g., comprehension, memory selection; Arthur, Bell, Edwards, Day, Tubre &

Tubre, 2005). It could be argued that work drivers are even more likely to under report

crashes and traffic violations during working hours due to fear of negative

repercussions from their organisations. Archival records of crash involvement can be

5 This book chapter is not part of the current program of research but was written during my work as a research assistant for Darren Wishart (external supervisor).


difficult to obtain, may breach anonymity and can be resource intensive (Arthur et al.,

2005). Some organisations may also be unwilling to share crash data to researchers.

These events are also relatively infrequent and, most times, the information gathered

from these events is retrospective and sensitive (Newnam et al., 2014). Therefore, there

should be a focus beyond crashes and traffic violations as measures of safety

performance within the work driving field.

Yet, researchers and practitioners within the general OHS domain are already

placing emphasis on concepts such as proactive safety behaviour to progress beyond

these traditional models of safety and to further improve safety at work. Instead of

focusing on lagging indicators of safety, OHS researchers are using proactive safety

behaviours as a measure of safety performance and as a leading safety indicator in high

risk occupations (Curcuruto & Griffin, 2017). Proactive safety behaviours are usually

described in contrast with safety compliance. Safety compliance refers to the

compulsory behaviours that individuals must perform in order to meet the minimum

requirements of safety in the workplace (e.g., wearing protective clothing and adhering

to safety procedures).

In contrast, proactive safety behaviour refers to behaviours that go beyond these

minimal safety requirements. These behaviours may not directly contribute to the

worker’s personal safety, but instead assist in developing an environment that supports

safety (Griffin & Neal, 2000). A way to have a proactive role in safety is to engage in

proactive safety behaviours, such as voicing concerns when safety issues arise, giving

suggestions during meetings, correcting unsafe procedures and reporting unsafe and

risky situations to management. These actions are also usually voluntary in nature

(Griffin & Neal, 2000).

It is important to acknowledge that while organisations have the legal duty to

provide a safe and healthy work environment for their employees, workers must also

actively contribute to the management of their workplace’s health and safety issues

(European Agency for Safety and Health at Work, 2012; Safe Work Australia, 2016b).

Various legislations on occupational and health safety in industrialised nations such as

Europe, United States and Australia require employers to consult and involve their

workers when addressing issues relating to the workplace’s health and safety. In turn,

workers are required to proactively participate and cooperate with their employers in

the management of their own and their co-workers’ safety while at work (e.g., Bureau


of Labor Statistics, 2014; European Agency for Safety and Health at Work, 2012;

Queensland Government, 2011). Within OHS legislations and governing bodies (such

as European Agency for Safety and Health at Work and Safe Work Australia), worker

participation often takes the form of safety consultations and formation of safety

committee and unions to represent workers regarding safety matters (Arrigo & Casale,

2010).

However, occupational health and safety researchers argued that workers could

proactively participate in resolving safety issues and improving their workplace health

and safety without such formal requirements (Curcuruto & Griffin, 2017).

Furthermore, these legislations usually require worker representatives to carry out

these consultations (European Agency for Safety and Health at Work, 2012). These

consultations and other representative tasks require additional hours to complete

besides their regular job duties. Therefore, support is usually needed, not only from

their employers, but from the workers they represent (European Agency for Safety and

Health at Work, 2012).

Curcuruto and Griffin (2017) argued that workers need to have a proactive role

in safety promotion, and they must work together with their co-workers and managers

in order to make meaningful improvements in workplace safety. Proactive safety

behaviours have been associated with reduced workplace accidents, injuries, work

accidents, lost work days and better reporting attitudes (e.g., Griffin & Neal, 2000;

Neal & Griffin, 2006). Research has also found that engaging in proactive safety

behaviour is more effective at improving safety outcomes in the longer-term

(measured by reduced accidents and workplace injuries) compared to merely

complying with safety rules and regulations (e.g., Christian et al., 2009; Neal &

Griffin, 2006; Neal et al., 2000). In Christian et al.’s (2009) meta-analysis on work

safety performance, they found that participating in safety activities has a slightly

stronger correlation and tighter confident intervals with reduced accidents and injuries

(Mp = -.15; 95% CI [-.18, -.13])6 compared to safety compliance (Mp = -.14; 95% CI

[-.20, -.08]). Furthermore, proactive safety behaviour has also been associated with

non-safety related outcomes such as better customer service and improved employees’

health (e.g., reduced blood pressure; Haslam, O’Hara, Kazi, Twumasi, & Haslam,

6 Mp = mean corrected pooled correlation


2016). These studies suggest that, for organisations to thrive and be successful in

achieving the desired high levels workplace safety, workers and management have to

work proactively towards risk reduction, transcending mere compliance with safety

standards and procedures (Curcuruto & Griffin, 2017).

Following this view, there is some merit in investigating proactive safety

behaviour as a measure of safety performance when employees drive for work. This

research aim is driven by the premise that, to improve safety while driving for work,

employees and management need to be proactive in addressing safety issues as

opposed to just simply following rules and regulations and reacting to crashes,

violations and injuries. Proactive safety behaviour as a measure of safety performance

in the work driving context could provide an alternative or complementary paradigm

to current risk management strategies. If work has less routine and predictability, such

as driving for work (due to other drivers on the road, road conditions, weather, etc.),

proactive safety behaviours may be required to guide safe practices where compliance

to rules and procedures may fall short (Zohar, 2008).

1.5 CHAPTER SUMMARY AND KEY LEARNINGS

This chapter summarised the literature on the issues of work driving and the

current practices and research on work driving safety management. Injuries and

fatalities resulting from work driving incidents is a significant public health issue. Due

to the costs of human lives and economic loss. However, current research on work

driving safety mostly focus on lagging indicators. Furthermore, management of risks

is often reactive, only conducting safety checks once an incident has already occurred.

While government regulations in Australia and other industrialised countries require

organisations to ensure their employees’ safety while driving for work, vehicle-related

incidents are still the cause of most work-related fatalities and injuries. This chapter

also highlighted the need for an alternative framework on work driving safety

management, with a particular focus on using HRO theory and Resilience Engineering

as theoretical lens to guide how work driving safety could be managed. The chapter

concluded with an appeal for a proactive measure of safety performance within the

work driving context, taking insights from research on proactive safety behaviours

from the OHS field.

16 Chapter 2: Proactive Safety Behaviours – A Review on Origins and Definitions

Chapter 2: Proactive Safety Behaviours – A Review on Origins and Definitions


A broad range of behaviours, concepts and terminologies is associated with the

construct of proactive safety behaviours. Thus, a comprehensive investigation of

proactive safety behaviours is required to gain a full understanding of this construct

before applying it to the work driving context. This chapter assists in placing the

foundation for the current program of research and critical to the background

information necessary for the subsequent chapters.

It is important to note that this taxonomy of behaviours has not been fully

examined within the work driving context and, therefore, most of the literature

reviewed within this chapter is derived from the areas of OHS and organisational

psychology. In particular, concepts of contextual performance, organisational

citizenship behaviours and work proactivity were reviewed in this chapter. Similar

concepts of these behaviours within the traffic safety research are also discussed.

2.2 BACKGROUND AND ORIGINS OF PROACTIVE SAFETY BEHAVIOUR

According to Griffin and Neal (2000), safety performance has two behavioural

components: safety compliance and safety participation. Safety compliance refers to

the compulsory behaviours that individuals must perform in order to meet the

minimum requirements of safety in the workplace (e.g., wearing protective clothing

and adhering to safety procedures). On the other hand, safety participation refers to

behaviours that go beyond these minimal safety requirements. These behaviours may

not directly contribute to the worker’s personal safety, but instead assist in developing

an environment that supports safety (Griffin & Neal, 2000). Safety participation is

voluntary in nature and usually involves helping co-workers, promoting safety

programs within the workplace, demonstrating initiative, and putting effort into

improving safety in the workplace (Griffin & Neal, 2000).

Chapter 2: Proactive Safety Behaviours – A Review on Origins and Definitions 17

Griffin and Neal’s model was an important study within the OHS field as it

provided a theoretical framework on safety performance, and identified its potential

organisational and individual antecedents (e.g., safety climate) and outcomes (e.g.,

injuries, accidents). Their seminal work also extended the dialogue on safety

performance as a multi-dimensional construct (Burke, Sarpy, Tesluk, & Smith-Crowe,

2002). Burke et al. (2002) argued that, prior to Griffin and Neal’s model, the literature

on OHS was mostly focused on safety compliance. In the past, interventions were

focused on increasing compliance with safety regulations and there was a common

assumption that compliance problems were the source of poor safety attitudes (Neal &

Griffin, 2002). Griffin and Neal’s (2000) research encouraged researchers and

practitioners to consider behaviours that go beyond safety compliance, and identify the

antecedents and determinants of such behaviours.

It is important to note, however, that while Griffin and Neal (2000)

conceptualised their model of safety performance, other researchers had already

examined safety behaviours that go beyond the minimal safety requirements. One of

the first studies that formally recorded such behaviours was conducted by Andriessen

(1978). While studying the safety behaviours of Dutch construction workers,

Andriessen (1978) noticed that, in addition to complying with safety rules and

regulations, construction workers also engaged in voluntary behaviours that improved

the safety of their work environment (e.g., voicing safety concerns to their

supervisors). Andriessen (1978) called these behaviours, ‘safety initiatives’.

Other notable early studies include Simard and Marchand’s (1995) investigation

of safety initiative behaviours within the manufacturing industry, and Cree and

Kelloway’s (1997) examination of willingness to participate in safety programs using

a sample of production employees. In both studies, safety initiatives and willingness

to participate in safety programs referred to workers’ readiness to engage in behaviours

that aimed to improve the safety of their working environment beyond the minimal

requirements of their jobs. Similarly, Roberts and Geller (2001) also proposed the

actively caring hypothesis, which refers to the “ultimate goal in occupational safety -

that employees care enough about the safety of their co-workers to act accordingly”

(p. 97). Workers who exhibit active caring behaviours are constantly looking for

organisational hazards, and when faced with unsafe behaviours and practices, these

workers intervene with corrective procedures (Roberts & Geller, 2001).


However, Griffin and Neal (2000) were the first to conduct a comprehensive

investigation of safety performance and to provide a model that differentiated between

compliant and participative safety behaviours, as well as to conduct an investigation

regarding their antecedents and possible safety outcomes. The current chapter utilises

Griffin and Neal’s (2000) research on safety participation as a starting point to

investigate the construct of proactive safety behaviour. Figure 2.1 provides the

overview of research on proactive safety behaviours, its related constructs and its

origins. To fully gain a comprehensive understanding of proactive safety behaviours

and its related constructs, it is imperative to discuss their origins, which are usually

derived from research in organisational psychology. These constructs are further

discussed in the following sections.

2.2.1 Contextual Performance

Safety participation as a component of safety performance is based on Borman

and Motowidlo’s (1993) theory of work performance. In this original theory, Borman

and Motowidlo (1993) proposed two major components of performance: task

performance and contextual performance. Task performance describes the extent to

which a worker effectively contributes to the organisation’s operations and processes

by directly providing the required services and products (e.g., firefighters performing

rescue operations). On the other hand, contextual performance describes the extent to

which a worker contributes to the organisation’s success by influencing the broader

psychological and social environment (i.e., context) in which the core tasks must

operate (e.g., firefighters picking up a shift when their co-workers are unwell; Borman

& Motowidlo, 1997). In this theory, contextual performance involves activities

Figure 2.1 Overview of the historical origins of proactive safety research

Organisational Citizenship Behaviour

(Organ, 1988)

Contextual Performance (Borman and Motowidlo,

1997)

Proactive Behaviours (Bateman and Crant, 1993)

Safety Citizenship Behaviours

(Hofmann et al., 2003)

Safety Participation (Griffin and Neal,

2000)

Proactive Safety Behaviours

(Curcuruto et al., 2015)

Prosocial Safety Behaviours

(Curcuruto et al., 2015)


directed at maintaining a work environment that allows the core task activities to be

effective. The taxonomy of contextual performance includes five categories of

behaviours (Borman & Motowidlo, 1997, p. 102):

1) persisting with extra effort and enthusiasm to successfully complete task

activities (e.g., perseverance and conscientiousness);

2) volunteering to carry out activities that are not formally part of their job

(e.g., suggesting organisational improvements);

3) helping and cooperating with others (e.g., assisting or helping co-workers);

4) following rules and procedures (e.g, meeting deadlines); and

5) endorsing, supporting, and defending organisational objectives (e.g.,

organisational loyalty).

Borman and Motowidlo (1997) argued that contextual performance is, in some

ways, more important than task performance especially when increased competition

and economic challenges (e.g., downsizing) demand companies to perform better.

Team members’ adaptability and willingness to exhibit extra effort contribute to the

success of the organisation in the longer term.

In a similar manner, safety participation describes behaviours that may not

necessarily directly contribute to workplace safety, but that may help develop an

environment or ‘context’ that supports safety (Griffin & Neal, 2000). Using the

conceptualisation of contextual performance, Neal and Griffin (2000) argued that

behaviours such as attending safety meetings and engaging in voluntary safety

activities promote a work context that is supportive of safety.

2.2.2 Organisational Citizenship Behaviour

Borman and Motowidlo’s (1993) operationalisation of contextual performance

is derived from the organisational citizenship behaviour research and, by extension,

safety participation shares several theoretical elements with organisational citizenship

behaviour as well. Organisational citizenship behaviour is defined as, “behaviours that

are discretionary, not directly or explicitly recognised by the formal reward system,

and that in the aggregate promote the effective functioning of the organisation”

(Organ, 1988, p. 4). Examples of organisational citizenship behaviours include helping

a co-worker finish a task and participating in various professional development

programs to benefit the organisation. Research shows that, although these behaviours

are not formally defined in one’s role, they are crucial to the survival, performance and


effectiveness of the organisation (Organ, Podsakoff, & MacKenzie, 2006). Employees

exhibiting high levels of organisational citizenship behaviours are often referred to as

‘good corporate citizens’ or ‘good soldiers’, in part due to the extra productive

activities undertaken for the advancement of the organisation and for the extra effort

beyond what is normally acknowledged, expected or rewarded (Organ, 1988).

Similar to Borman and Motowidlo’s (1993) taxonomy of behaviours that relates

to contextual performance, Organ’s (1988) original research on organisational

citizenship behaviour also had five dimensions: altruism, conscientiousness,

sportsmanship, courtesy and civic virtue behaviour. More recent research, however,

has indicated seven dimensions, which expanded and revised the taxonomy of

behaviours related to organisational citizenship behaviours to include a more

comprehensive taxonomy of behaviours (Organ et al., 2006), and which are: helping,

individual initiative, organisational loyalty, self-development and organisational

compliance. Sportsmanship and civic virtues were the only two dimensions that

remained from the original research.

Using the dimensions studied within the organisational citizenship research,

Hofmann, Morgeson, and Gerras (2003) coined the term ‘safety citizenship

behaviours’ to describe the extra-role behaviours that workers carry out to improve the

safety of their work environment. Hofmann et al.’s (2003) research on safety

citizenship behaviours expanded the construct of safety participation. Safety

participation is unidimensional in nature, and is usually measured with these three

behaviours: “I promote the safety program within the organization”, “I put in extra

effort to improve the safety of the workplace” and “I voluntarily carry out tasks or

activities that help to improve workplace safety” (Neal and Griffin, 2006, p. 953). On

the other hand, Hofmann et al.’s (2003) research on safety citizenship behaviours

provided a higher-order construct consisting of six dimensions of proactive safety

behaviours. These six dimensions are derived from the organisational citizenship

behaviour research, and include (Hofmann et al., 2003, p. 178):

Helping – behaviours that help out the organisation and its members to ensure

that everyone is working safely

Safety Civic Virtue (Keeping Informed) – behaviours that are related to

maintaining an up-to-date knowledge on the safety issues in the organisation.


Initiating Safety-Related Change (Improving Safety) – behaviours that try to

improve the overall safety of the organisation that are self-initiated

Voice – voicing out safety-related recommendations and opinions on safety

matters even if others may disagree

Stewardship – taking action to prevent safety violations and protect the well-

being of their co-workers

Whistleblowing – similar to the voice dimension but behaviours that are

specific to reporting and monitoring of other members’ unsafe behaviours and

potential safety violations

Workers who exhibit safety citizenship behaviours tend to be more compliant

with safety management policies and are involved in fewer work related safety

incidents (Hofmann et al., 2003). Hofmann et al. (2003) suggested that safety

citizenship behaviours are associated with the promotion and improvement of safety

within the workplace. Such behaviours can enhance individual commitment and

personal responsibility, which may translate into improvements in tangible risk

management and consequently improved safety outcomes (Wishart et al., 2019).

2.2.3 Proactivity

Another source of research when investigating proactive safety behaviours is the

literature on proactivity (e.g., Crant, 2000; Parker, Williams, & Turner, 2006).

Research on work proactivity began with the construct of proactive personality, where

Bateman and Crant (1993) argued that some individuals have the disposition to initiate

change in their immediate work environment, while some individuals have no self-

initiative and are merely reacting to their work situations. However, recent studies in

the work proactivity literature place emphasis on the behaviour aspect of proactivity

instead of the dispositional characteristics (e.g., Crant, 2000; Parker et al., 2006). By

conceptualising work proactivity as a behaviour, proactivity is seen as something that

leaders can facilitate and shape within the organisation, rather than seeing it as a fixed

individual attribute that someone may or may not have (Crant, 2000). This way, leaders

are encouraged to promote and shape proactive behaviours among the workforce.

Proactive behaviours in the workplace are conceptualised as self-stating, change-

oriented and future-focused. These elements suggest that proactive behaviours are self-

initiated rather than enforced (Parker et al., 2006). Employees who exhibit high levels


of proactive behaviours enact and drive change within the organisation rather than

adapting to their situation (Parker & Collins, 2010). Being future-focused suggests that

proactive behaviours involves anticipating and thinking ahead, rather than reacting to

the current situation (Grant & Ashford, 2008; Parker & Collins, 2010).

Similar to Borman and Motowidlo’s (1997) research on contextual performance,

research on proactivity took concepts from the organisational citizenship behaviour

literature. However, while there are similarities between organisational citizenship

behaviour and proactive behaviours, researchers have argued for differences between

the two constructs. Parker, Bindl, and Strauss (2010) contested that proactivity is

different from organisational citizenship behaviour because proactivity can be

conceptualised as a stable personality trait, a behaviour or a process, whereas

organisational citizenship behaviours only describes extra-role behaviours in the

workplace. For instance, an employee may carry out an in-role behaviour

‘proactively’. Additionally, there are some behaviours under the organisational

citizenship behaviour construct which can be seen as passive or compliant (e.g.,

following rules and being loyal), instead of being ‘proactive’ (Belschak & Hartog,

2010). Therefore, Parker et al. (2010) argued that the proactivity literature investigates

the active, taking-charge and change-oriented aspect of citizenship behaviours.

Recently, research by Curcuruto and colleagues have incorporated the

proactivity literature in their investigation of proactive safety behaviours (Curcuruto,

Conchie, Mariani, & Violante, 2015; Curcuruto, Mearns, & Mariani, 2016). Curcuruto

et al. (2015) suggested that these previous constructs of proactive safety behaviour,

safety participation and safety citizenship could be distinguished into two categories:

proactive and prosocial behaviours. Curcuruto et al. (2015) argued that proactive

safety behaviours should be action-oriented, change-oriented and future-focused, as

opposed to passively reacting to the events that have already occurred and simply

complying with rules and regulations. Prosocial safety behaviours, on the other hand,

are affiliative in nature, and mostly reflect the helping and stewardship dimensions of

organisational citizenship behaviours (Curcuruto et al., 2015). With this

conceptualisation, Curcuruto et al. (2015) argued that the dimensions of whistle-

blowing, safety voice, safety initiative and keeping-informed identified within the

safety citizenship research are proactive safety behaviours, while helping and

stewardship dimensions are prosocial safety behaviours.


However, it could be argued that helping co-workers be safe and protecting

colleagues from hazards still have aspects of proactivity given that these behaviours

are usually discretionary and not expected from one’s roles. Furthermore, protecting

co-workers from hazards require an initiative to act. Therefore, the current program of

research argues that prosocial behaviours (e.g., stewardship, helping) are still change-

oriented and action-focused behaviours that seek out to bring positive change in safety

practices within organisations through the process of fostering co-operation and social

relationships with co-workers. Consequently, these behaviours are still regarded as an

aspect of safety proactivity.

2.2.4 Bringing it all together

Due to the different concepts that inform the research on proactive safety

behaviours, various terminologies and definitions of proactive safety behaviours have

been used interchangeably within the literature (see Table 2.1).

Table 2.1 Different terminologies associated with the construct of Proactive Safety Behaviours adapted and extended from Curcuruto and Griffin (2017, pp 109-110).

Terminology / Construct Definition Behaviours measured Safety Initiative (Andressien, 1997; Simard & Marchard, 1995)

Employees’ attempts to create a safer work environment.

Employees take initiatives in order to work more safely Make suggestions in order to improve safety at work environment Put pressure on the supervisor to improve safety at work

Actively caring about safety (Geller, Johnson, Redmon, & Mawhinney, 2001; Geller, Roberts, & Gilmore, 1996)

Employees acting to benefit the safety of other employees.

Observing and recording the safe and unsafe behaviours of coworkers, and then provide constructive behavioral feedback

Willingness to participate in health and management programs (Cree & Kelloway, 1997)

By participating in health and safety programs, employees have the opportunity to have a "voice" in the management of workplace health and safety.

Participation in various programs (e.g., attending safety meetings, attending safety training).

Safety Participation (Griffin & Neal, 2000)

Safety behaviours that may not necessarily directly contribute to workplace safety, but help develop an environment that supports safety.

Promoting safety program Extra effort to improve safety Voluntarily carry out tasks to improve safety

Safety Citizenship (Hofmann et al., 2003)

Safety behaviours that are not formally rewarded by the organisation, but plays a strong role in accident prevention due to its facilitation in improving the safety and learning within the workplace.

Whistle-blowing Safety voice Helping Safety initiative


Proactive safety behaviours (Curcuruto et al., 2015)

Safety behaviours that are challenging in nature and seek to bring out positive change in the safety practices within the organisation.

Whistle-blowing Safety voice Safety initiative Keeping informed

Prosocial safety behaviours (Curcuruto et al., 2015)

Safety behaviours that are affiliative in nature. This behaviour generally manifest itself as helping co-workers and looking out for their welfare and safety. Often seen as a way to foster good relationships.

Helping Stewardship

Using the origins of contextual performance, organisational citizenship

behaviour and proactivity research, the overarching conceptualisation of proactive

safety behaviours that is utilised in the current program of research has the following

distinct, yet related, components:

1) behaviours that improve the context of the work environment to be more

supportive of safety – these behaviours may not directly contribute to

workplace safety, but may facilitate an environment that supports safety

(i.e., behaviours that create a positive environment for safety);

2) behaviours that aim to improve workplace safety that cannot be forced (i.e.,

self-starting); and

3) behaviours that are change-oriented and which aim to improve the current

workplace safety practices.

Taking the current knowledge on proactive safety behaviours, a more specific

definition of proactive safety behaviours within the work driving safety context could

be defined as: behaviours that work drivers perform that create a work environment

that promotes work driving safety. These behaviours must be 1) self-starting (in other

words, cannot be forced), and/or 2) change-oriented (with an aim to improve the

current work driving safety practices). Coming from the perspectives of safety

citizenship and safety proactivity research, arguably this construct is multi-

dimensional, which may include: voicing out, intervening, problem prevention and

changing the organisation's policies and procedures.

2.3 SIMILAR CONCEPTS IN TRAFFIC SAFETY RESEARCH

While the concept of proactive safety behaviours had not been studied in the

work driving context, similar constructs have been investigated in the general road

safety research. For instance, prosocial driving behaviours have been studied in the


literature. Prosocial driving has been referred to as, “a pattern of safe driving

behaviours that potentially protect the well-being of passengers, other drivers, and

pedestrians, and that promotes effective cooperation with others in the driving

environment” (Harris et al., 2014, p. 2). Ross and Antonowicz (2004) argued that most

people characterised driving as an individual activity, whereas in reality, driving

occurs in a ‘social vacuum’. The social feedback drivers receive from other road users

can, implicitly and explicitly, influence the driver’s decision-making and behaviour on

the road. Being a safe and courteous driver requires adequately developed social

cognitive skills and positive values that show concern to the safety of oneself and

others (Ross & Antonowicz, 2004).

However, prosocial driving behaviours describe general driving behaviours on

the road (e.g., “decreasing speed to accommodate poor road conditions”, p. 3) rather

than specific behaviours within the work context. Furthermore, prosocial driving

behaviours often reflect compliance with traffic laws (e.g., “obey traffic signs” and

“come to a complete stop at a stop sign”; Harris et al., 2014, p. 3) instead of personal

initiatives and change-oriented behaviours.

The concept of traffic safety citizenship was also developed by Ward and

colleagues (Otto, Finley, & Ward, 2016; Finley, Riggs, Otto, & Ward, 2015). Based

on the concept of safety citizenship behaviours, traffic safety citizenship focused on

the prosocial aspects of traffic behaviours across different levels of society (e.g.,

individual family, workplace and community). Examples of such behaviours include:

asking others not to speed, intervening to prevent others from drink driving and

advocating for stronger traffic laws. This concept is based on the premise that, often,

it is only a small number of public road users who engage in risky behaviours.

Therefore, it is important to foster more active engagement from the larger majority of

safe road users to influence the behaviours of the risky drivers.

In a survey of 1,954 adult drivers in the United States, Otto et al. (2016) found

that around half of the respondents indicated being in a situation in the past 12 months

when a passenger was not wearing a seat belt or a driver was using a mobile phone

while driving. Participants were more likely to intervene if the person who conducted

the unsafe behaviour was part of their close social circle (e.g., family or friends)

compared to more socially distant circle (e.g., acquaintances or strangers).

Furthermore, Otto et al. (2016) found that the participants were most likely to report


intervening if they perceived that other people would also intervene. When group

differences were assessed between individuals who reported higher frequency of

intervening compared to those who rarely or never intervened, participants who

intervened more were more likely to have a higher sense of comfort and confidence in

performing such behaviours.

A similar study on these intervening behaviours were demonstrated in a study

by Buckley, Sheehan and Chapman (2009) on a school-based behaviour change

program that targets friendship groups for risk driving prevention. In their study, they

examined the idea of ‘mateship’ in encouraging adolescents to intervene when their

friends engage (or are thinking of engaging) in risky driving behaviours such as drink

driving and speeding (Buckley et al., 2009). Another school-based road safety program

by Kennedy, Cullen, Firman, Fleiter and Lewis (2018) examined how the program

could influence young people’s intention to speak up when a driver engages in a risky

behaviour. These interventions include voicing concerns, taking away their friend’s

keys or offering a lift to prevent their friends from engaging in risky driving

behaviours.

This concept of engaging bystanders to reduce risky driving behaviours were

also examined within the workplace context. Otto, Ward, Swinford and Linkenbach

(2014) studied employees’ intention to prevent their co-workers from engaging in

risky driving behaviours (e.g., not wearing a seatbelt, texting while driving, using a

mobile phone while driving, aggressive driving or speeding). Their findings revealed

that, individuals with higher intentions to intervene were associated with stronger

perceived norms that intervening behaviours were supported or approved of within the

workplace.

These intervening behaviours tap into the concepts of safety stewardship

dimension of safety citizenship behaviours. While these studies present a starting point

for the current program of research, there may be other proactive safety behaviours

that employees could engage in while driving for work. More specifically, these

studies (e.g., Buckley, Sheehan, & Chapman, 2009; Kennedy et al., 2018; Otto et al.,

2016; Otto et al., 2014; Ward, 2015) only examined the aspect of intervening when

someone is engaging in a risky driving behaviours. Other aspects of proactive safety

behaviours (e.g., safety voice, safety civic virtue) were not yet investigated within the

context of work-related driving. Therefore, a significant gap exists within the research


of proactive safety behaviours and in terms of understanding its nature and

multidimensionality within the context of work driving safety. Due to the specific

context of work driving, there may be other factors that previous research has not

examined, that could influence one’s engagement with proactive safety behaviours.

The current program of research sought to address this gap.

2.4 CHAPTER SUMMARY AND KEY LEARNINGS

This chapter has critically reviewed and discussed the origins and definitions of

proactive safety behaviours using studies from organisational psychology and general

occupational safety research (Figure 2.1). The chapter also examined similar

constructs within traffic safety research. The review informed the definition of

proactive safety behaviours used by the current research program. This definition

comprises three key components: behaviours that may not directly contribute to

workplace safety but they facilitate an environment that supports safety (in other

words, behaviours that create a positive environment for safety); behaviours that aim

to improve workplace safety that cannot be forced (self-starting); and, behaviours that

are change-oriented which aim to improve the current workplace safety practices

(change-oriented behaviours). The literature reviewed in this chapter helped to identify

the gaps in the literature and to inform the background research that was necessary to

develop the research aims, objectives and questions of the current program of research.

Figure 2.2 Fields of research that helped inform the proactive safety behaviour construct within the

work driving context and the foundation of the current program of research

Proactive Safety Behaviours within the work driving context

Occupational Health and Safety research

Organisational psychology research

Road traffic safety research

Chapter 3: Overview of Research Program 29

Chapter 3: Overview of Research Program


The work driving context is complex and dynamic, involving different entities

and decision makers (Stuckey et al., 2010). Not only do work drivers have to safely

operate within the road environment (which is typically the domain of road

authorities), they also have to perform within the occupational context (Stuckey et al.,

2010). However, work driving safety management typically focuses on on-road

context, with little to no regard on the influence that organisations could do to manage

risks related to work driving. Driving for work also has unusual characteristics that

may affect safety management such as irregular work driving times, unrealistic work

demands, driving seen as secondary or not part of one’s job and may involve remote

work without direct supervision (Newnam, Greenslade, Newton, & Watson, 2011).

Within the general OHS field, a substantial amount of work has been conducted

on the development of safety management systems that have yet to be applied within

the work driving context (Mooren, 2016). Some risk management models

acknowledge the dynamic and multi-dimensional nature of work and increasingly

recognises the complex processes involved in managing risks. Some approaches

require organisations to continually anticipate risks and be responsive and proactive in

eliminating risks. The nature of road transport tasks involves exposure to a relatively

uncontrolled work environment (e.g., public roads). It is therefore important for work

drivers and managers to continually anticipate risks and adapt the process to avoid

injury crashes (Newnam, Sheppard, et al., 2014). Currently, there is little research on

this proactive approach to safety management within the work driving context.

However, current studies in other occupational settings could provide some foundation

research in the necessary management and organisational characteristics that might

influence safety performance. Newnam et al. (2014) aptly described the need for

proactive safety behaviours in the work driving setting with this statement:

“To achieve reductions in work-related road traffic injury, it will be necessary

to extend the focus beyond the individual’s compliance with safety procedures, as safe

30 Chapter 3: Overview of Research Program

work environments depend on individuals anticipating threats to safety, showing

concern for the safety of others (team supervisor) and contributing to safety” (p. 4)

The previous two chapters discussed the background and rationale of the

program of research, focusing on the issues of work driving and the current

management and research on work driving risks. The previous chapters also identified

the gaps in research and presented the current literature on proactive safety behaviours.

The current chapter is an extension of the previous chapters and provides an overview

of the research program. The current chapter presents the research aim and objectives

as well as the demarcation of scope that describes the context and constraints in which

the program of research were conducted. Finally, the structure of the thesis document

is also presented.

3.2 RESEARCH QUESTIONS, AIM AND OBJECTIVES

While proactive safety behaviours have been studied in various work settings,

proactive safety behaviours have not been fully examined within the work driving

context. This presents a significant gap in the current work driving literature.

Therefore, the following questions had been developed and formed the foundation of

the current program of research:

RQ1: What do we currently know about the construct of proactive safety

behaviours?

RQ2: What are the proactive safety behaviours that work drivers perform to

ensure and improve safety (and that of their co-workers) while driving for

work?

RQ3: What are the individual and organisational factors that promote and

inhibit proactive safety behaviours in work drivers?

Relating to these questions, this research program has an overarching aim, with

three-related objectives:

Overarching Research Aim: To develop a research model and measure of proactive

safety behaviour that could be utilised within the context of work driving.

Research objective 1: To investigate the current literature in order to better

understand the proactive safety behaviour construct and identify the factors that

influence employees’ engagement in proactive safety behaviours. This


investigation will inform the development of the research model and

measurement.

Research objective 2: To develop an instrument to measure proactive safety

behaviours within the work driving context.

Research objective 3: To investigate the suitability of the proposed model by

conducting a quantitative study in a sample of work drivers.

3.3 DEMARCATION OF SCOPE

Work drivers in this program of research were defined as employees who drive

as a requirement in order to perform important aspects of their jobs. The target

population were required to drive for work at least once a week, including the commute

to and from work7. The target population spans a range of industries including remote

and mobile workers who travel to different locations throughout the day (e.g., utility

workers, teleworkers, community nurses, emergency services and couriers; Huang et

al., 2013; Newnam & Watson, 2011b).

The context of work driving offers a unique perspective on the study of

occupational safety performance. Work drivers could range from having traditional

(e.g., office desk job) to precarious (e.g., field work) employment arrangements. In

addition, their work involves the use of the public road system, thus operating within

both occupational health and safety and road safety policy contexts (Fort et al., 2016).

Where some workers may only have to comply with occupational health and safety

policies specific to their jobs, driving for work also requires compliance with traffic

laws and regulations. Work drivers are also usually on the road alone without

supervision, making it is more difficult to monitor their behaviour compared with

employees within a single physical setting (e.g., factory workers; Huang et al., 2013;

Newnam, Lewis & Watson, 2012). Supervisor absence could also mean increased

reliance on themselves to drive safely for work (Huang et al., 2013; Newnam et al.,

7 Although some researchers regarded commuting to and from work as distinct from work-related travel (Fort, Ndagire, Gadegbeku, Hours, & Charbotel, 2016; Health and Safety Executive, 2014; Salminen, 2000), it is important to acknowledge that commuting to and from work is considered as ‘work-related’ travel within the work health and safety legislation and work insurance claims (especially within Queensland; WorkCover Queensland, 2016). Other researchers have also argued that commuting to and from work should be considered a work-related driving activity (e.g., Newnam & Watson, 2011a, 2011b).

32 Chapter 3: Overview of Research Program

2012). Therefore, there are specific factors within the work driving context that could

influence one’s engagement with proactive safety behaviours.

3.4 THESIS OUTLINE

This thesis has four chapters that document the program of research together

with two introductory chapters, and one general discussion and concluding chapter

(see Figure 3.1 for the thesis outline). Chapter 1 provided the introductory discourse

on the research topic, outlining the background and rationale of the program of

research. Chapter 1 specifically discussed the current research on work driving safety,

detailing the issues of work driving, the shortcomings of current work driving safety

management and the need for an alternative framework for managing risks related to

work driving safety. Chapter 1 also presented the theoretical framework that guided

the the current program of research. Chapter 2 discussed the topic of proactive safety

behaviours, its background and origins on organisational psychology, its application

on the OHS field and its possible role in traffic psychology and specifically within

work driving context. The current chapter (Chapter 3) provides an overview of the

research context, questions and aim of the research program. Chapters 4 to 6 document

the three studies that constitute the program of research. Chapter 7 discusses the overall

findings of the program of research and concluding statements.


Figure 3.1 Outline of the program of research and its presentation in this thesis

Research Question 1: What do we currently know about the construct of proactive safety behaviours? Related research objective: To investigate the current literature in order to better understand the proactive safety behaviour construct and identify the factors that influence employees’ engagement in proactive safety behaviours. (RO1) Method: Literature review and meta-analysis

Overarching Research Aim: To develop a new model and measure of proactive safety behaviours within the work driving context.

Chapter 2 – Proactive Safety Behaviours – A Review on Origins and Definitions (Literature review)

Chapter 4 – Meta-analytic Review of Antecedents to Proactive Safety Behaviours (Study 1)

Research Question 2: What are the proactive safety behaviours that work drivers perform to improve safety (and of their co-workers) while driving for work? Related research objective: To develop an instrument to measure proactive safety behaviours within the work driving context. (RO2) Method: Item pool generation with the research team and expert panel for the scale development (Study 2a). Online survey with sample of work drivers to test the developed scale (Study 2b and Study 2c).

Study 2a – Item Generation and Expert Panel

Chapter 5 – Development and Testing of Proactive Safety Behaviour Measure within the Work Driving Context

Study 2b – Pilot Study

Study 2c – Main Survey

Research Question 3: What are the individual and organisational factors that promote and inhibit proactive safety behaviours in work drivers? Related research objective: To investigate the suitability of the proposed model by conducting a quantitative study in a sample of work drivers. (RO3) Method: Online survey with sample of work drivers.

Chapter 6 – Model Testing of Proactive Safety Behaviours within the Work Driving Context (Study 3)

34 Chapter 4: Meta-analytic Review of Contextual and Proximal Antecedents to Proactive Safety Behaviours

Chapter 4: Meta-analytic Review of Contextual and Proximal Antecedents to Proactive Safety Behaviours


Behaviours that go beyond one’s expected roles “do not exist in a vacuum”

(Gyekye & Salminen, 2005, p. 815). Organisational factors, individual differences,

cognitive mechanisms, and the dynamic relationship between these variables, promote

or inhibit proactive safety behaviours (Curcuruto & Griffin, 2017). Using Curcuruto

and Griffin’s (2017) framework, the current chapter appraises and synthesises studies

that have examined the distal and proximal antecedents of proactive safety behaviours.

Proximal antecedents refer to factors that are directly responsible for the differences

in proactive safety behaviours (Griffin & Neal, 2000). On the other hand, distal

antecedents (which may include variables within the work environment) may also

directly influence proactive safety behaviours, but the effect is usually smaller

compared to proximal antecedents. Distal antecedents may also influence proactive

safety behaviours via the proximal antecedents (Griffin & Neal, 2000). Griffin and

Neal (2000, p. 347) argued that distinguishing distal from proximal antecedents is

important in order to “allow a systematic assessment of conceptually distinct

perceptions that may have different causes and consequences within organisations”. In

other words, it is critical to identify these factors as it will facilitate a better

understanding on how management can encourage and promote these behaviours in

the workplace. This chapter forms the first study of the current program of research

(Study 1) and aims to answer RQ1: “What do we currently know about the construct

of proactive safety behaviours?”

4.1.1 Purpose of the Study

Compared to previous reviews that examined other safety behaviours (Christian

et al., 2009; Nahrgang, Morgeson, & Hofmann, 2011), the current review focused on

the antecedents of proactive safety behaviours and its associated constructs. The meta-

analytic component of the review also extends upon Curcuruto and Griffin’s (2017)

Chapter 4: Meta-analytic Review of Contextual and Proximal Antecedents to Proactive Safety Behaviours 35

literature review on safety proactivity. While Curcuruto and Griffin (2017) provided a

comprehensive account of safety proactivity, this current meta-analysis presents an

opportunity to quantify the relationships of the different contextual and proximal

variables that they proposed within the model. In addition, the meta-analytic

calculations allowed the pooling of available quantitative data from individual studies

for a more precise estimate of the effect size of each relationship (Schmidt & Hunter,

2014). The current review also informs the development of a theoretical model of

proactive safety behaviour that would be applied within the work driving context.

Therefore, general discussions of the antecedents in relation to work driving safety are

conducted, particularly if similar constructs had been studied within the work driving

safety field. Studies that were found during the database search that may relate to work

driving safety were also specifically examined as part of the meta-analysis.

4.1.2 Scope of the Study

While previous studies have examined the role of individual differences and

dispositional characteristics on proactive safety behaviours, the current review will not

examine these variables. Within the general safety literature, various studies have

investigated a vast array of personality characteristics that predispose individuals to

errors, accidents and injuries. The search for reliable individual difference variables

that strongly relate to safety performance is characterised by inconsistent and

inconclusive findings (Hogan & Foster, 2013). Several reviews that have examined

the role of personalities and dispositional characteristics often show weak or non-

significant effects on safety behaviours (Beus, Dhanani, & McCord, 2015; Christian

et al., 2009). In addition, this perspective is influenced by traditional models of safety

which are based on the premise that “error-prone” individuals contribute the most

towards workplace accidents and injuries (Hogan & Foster, 2013). Therefore, this

review only focused on the contextual and proximal antecedents of proactive safety

behaviours.


4.2 FRAMEWORK FOR MODELLING PROACTIVE SAFETY BEHAVIOURS

Despite the importance of proactive safety behaviours in improving workplace

health and safety, it is only recently that a conceptual model of safety proactivity has

been proposed. According to Curcuruto and Griffin (2017), aiming for an injury-free

workplace requires improvement in the work environment, the person and the

behaviour. These domains are interdependent and dynamic – a change in one domain

impacts on the others. Curcuruto and Griffin’s (2017) model proposes that there are

distal and proximal antecedents of proactive safety behaviours. Curcuruto and Griffin

(2017) argued that distal antecedents comprise contextual and individual factors that

impact on cognitive-motivational mechanisms (i.e., proximal antecedents) which, in

turn influence whether one performs a proactive safety behaviour (See Figure 4.1).

However, it is important to acknowledge that this framework is influenced by previous

models on safety performance by Griffin and Neal (2000) and work proactivity by

Parker and colleagues (Parker & Wang, 2015; Parker, Williams, & Turner, 2006).

While the specific factors studied within these three frameworks differ, there are

several notable commonalities. For instance, among these three frameworks, the

researchers have recognised the importance of dispositional characteristics that may

predispose individuals to act proactively, the role of work environment in cultivating

these behaviours and the potential cognitive mechanisms that have a direct impact on

these behaviours.

Contextual Antecedents (climate, support, job

design)

Individual Antecedents (competency, affective

states and role definition)

Motivational - Capability Drivers

(control perception; self-efficacy)

Motivational -

Commitment Drivers (psychological ownership;

felt responsibility)

Future Oriented States (prevention orientation;

improvement orientation)


Distal Antecedents Proximal Antecedents

Figure 4.1 Curcuruto and Griffin’s (2018) model of antecedents of safety proactivity. The grey shaded boxes were the focus of the current meta-analysis.


The current review utilised this framework for the purpose of organising both

the literature and the results of the meta-analysis. Within the review, the antecedents

were classified as either contextual (i.e., organisation-related) or proximal (i.e.,

cognitive-psychological mechanisms). Contextual antecedents were considered to be

more distal compared to proximal antecedents and, therefore, it was anticipated that

proximal antecedents would yield larger relationships with proactive safety behaviours

compared to the distal contextual factors.

4.2.1 Contextual Antecedents

To engage in proactive safety behaviours, workers must perceive that such

behaviours are valued and important within their work environment (Curcuruto &

Griffin, 2017; Griffin & Curcuruto, 2016; Griffin & Neal, 2000). In this review,

contextual antecedents refer to variables specific to the organisational environment

that have been found to affect employees’ proactive safety behaviours. In Curcuruto

and Griffin’s (2017) safety proactivity framework, the contextual antecedents were

obtained from the paradigms of: 1) organisational climate and culture, 2)

organisational support and social exchange theories, and 3) work design. Similar to

their framework, the current review presents three categories of contextual antecedents

of proactive safety behaviours: 1) safety climate, 2) leadership, perceived

organisational support and trust, and 3) work demands and work autonomy.

Safety Climate

Researchers often look at the organisation’s safety culture and/or safety climate

when investigating the organisational factors that influence safety behaviours in the

workplace. Safety climate represents the inferences that the employees make towards

their organisation’s safety practices (Zohar, 1980). Safety climate is also often used as

an interchangeable term for safety culture (Mearns & Flin, 1999). While there are some

conceptual differences between safety climate and safety culture, both concepts

represent the importance and value that the organisation and its members place on

workplace safety (Guldenmund, 2000)8.

8 The discussion on the conceptual differences between safety culture and safety climate is beyond the scope of this study. However, it is generally accepted that safety climate provides a ‘snapshot’ of safety culture, and that it reflects the ‘overt manifestations’ of safety culture, such as attitudes, policies and practices of safety within the organisation. Safety culture refers to the shared values, norms, beliefs and ideas of safety within an organisation (Guldenmund, 2000) and it is crucial in shaping and defining how


According to Zohar (2010), organisations have complex rules and policies across

competing domains. Employees must then sort through these complexities to discern

which behaviours are expected, supported and rewarded. How managers and

supervisors trade-off the need for productivity, while ensuring safety policies and

procedures are still being implemented, presents a message to employees on which

behaviours are most important. If productivity is favoured over safety, employees are

more likely to prioritise production and usually to the detriment of safety (Zohar,

2010). On the other hand, a positive safety climate prioritises safety over the

organisation’s productivity levels. Thus, a positive safety climate would encourage

safety compliant and enhancing behaviours especially if these behaviours are

supported and rewarded.

Safety climate is often seen as a meaningful antecedent to safety behaviour and,

in particular, proactive safety behaviours. The original conceptualisation and

operationalisation of safety climate was conducted at the organisational level (e.g.,

Zohar, 1980), meaning that the shared perceptions of safety practices are coming from

the policy and procedural actions of top management and practices of the supervisors

(Brondino, Silva, & Pasini, 2012). However, Zohar and Luria (2005) proposed that

employees examine their work environment from dual perspectives of being members

of both the organisation and of a particular sub-unit within that organisation. While

safety climate was originally conceptualised as established safety policies and

procedures from the top-level management, sub-unit managers and supervisors

execute these policies and procedures through direct interaction with subordinates

(Zohar & Luria, 2005). Thus, potential discrepancies may occur between the top-level

safety is presented and integrated in the organisation’s daily processes, functions and management. Safety culture is complex as it is comprises the underlying and enduring structure of the fundamental values, norms and assumptions towards safety in an organisation, some of which may not be observable by its own members (Mearns & Flin, 1999). Several researchers argued that qualitative and ethnographic methods should be used when investigating an organisation’s safety culture (Weigmann et al., 2002). The deeper structure of an organisation’s culture is often not immediately accessible to outsiders, therefore, in-depth and extensive measures should be used to gain a full understanding of an organisation’s safety culture (Weigmann et al., 2002).

Since safety climate is easier to measure using quantitative methods (Guldenmund, 2000), the majority of survey studies used the construct of safety climate. Further, the current study utilised meta-analysis to quantify the effects of the observed relationships, therefore, the review (and the current program of research) utilised the construct of safety climate to provide a ‘snapshot’ of an organisation's safety culture.


(i.e., organisation-level) established policies and procedures and the sub-unit (i.e.,

group-level) execution of these policies and procedures.

This multi-level analysis of safety climate only gained research attention within

the last 10 to 15 years and, previously, safety climate was primarily measured from an

individual perspective. Thus, distinction between organisational-level and group-level

safety climate has tended to be poorly specified (Zohar & Luria, 2005). While safety

climate is continuously being measured from the individual level, there is growing

evidence for the predictive and informative nature of safety climate at various levels

of aggregation (Griffin & Curcuruto, 2016; Zohar, 2008, 2014). For instance, Clarke’s

(2006) review showed that an overall measure of safety climate has a stronger effect

with safety participation (Mp = 0.50) compared to safety compliance (Mp = 0.43).

However, when Christian et al. (2009) distinguished the measurement of safety climate

between psychological (i.e., individual-level) and group-level, the differential

influence of the two levels were more pronounced – with individual-level safety

climate having a stronger relationship with safety compliance (Mp = 0.49) compared

with the relationship between group-level safety climate and safety compliance (Mp =

0.40). On the other hand, the effect size found between individual-level safety climate

and safety participation (Mp = 0.59) was the same with the effect size found between

group-level safety climate and safety participation (Mp = 0.59).

The current study primarily built on previous meta-analyses by studying the

effect of different levels of safety climate (e.g., individual, work group-level, and

organisational-level) on proactive safety behaviours. Clarke’s (2006) meta-analysis

did not distinguish between the different levels of safety climate. Christian et al.’s

(2009) meta-analysis only examined the psychological and group-levels of safety

climate on safety participation, but did not examine the impact on organisational-level

safety climate on safety participation9.

In the current study, the definitions used for individual and group-level safety

climate were based on Christian et al.’s (2009) review to maintain consistency and

9 Christian et al. (2009) only looked the relationship between organisational-level safety climate and safety performance (i.e., a combination of safety compliance and safety participation). This could be due to the small number of studies in their review that utilised the organisational-level analysis. The current review found enough studies for the organisational-level to enable that factor to be included in the analysis.


direct comparison to their results. Individual-level climate is defined as: “individual

perceptions of safety-related policies, practices, and procedures pertaining to safety

matters that affect personal well-being at work” (Christian et al., 2009, p. 1106). When

these perceptions are shared among individuals in a particular work group

environment, a group-level climate emerges. Therefore, group level safety climate is

defined as, “shared perceptions of work environment characteristics as they pertain to

safety matters that affect a group of individuals” (Christian et al., 2009, p. 1106). Since

Christian et al. (2009) did not specifically examine organisational-level safety climate

and no definitions were provided, Zohar et al.’s (2008) definition of organisational-

level safety climate was utilised: “shared perceptions among members of an

organisation with regard to its fundamental properties, i.e., policies, procedures, and

practices” (p. 376).

Relevance to work driving safety

While safety climate research within the work driving setting is limited,

available research suggests that safety climate has a role discouraging risky driving

behaviour in workers ( Newnam et al. 2008; Wills, Watson, & Biggs, 2006, 2009). For

instance, in a survey of 323 employees working in Australian fleet settings, Wills et

al. (2009) found that employee’s perception of safety climate was a significant

predictor of unsafe driving behaviours and driving intentions over and above

demographic characteristics (age and gender) and driving exposure (annual mileage

and weekly driving hours). Safety climate explained the largest amount of unique

variance in the driver’s reported behaviours compared to other factors (i.e., safety

attitudes, perceived behavioural control and driving experience; Wills et al., 2009).

These results provide evidence of the importance of having policies and rules

regarding safe driving and of the importance of safety climate on work driving safety.

While previous research has demonstrated the role of safety climate on driving

behaviours, it is possible that it may also have an influence on work drivers’ proactive

safety behaviours.

Leadership, Perceived Organisational Support and Organisational Trust

Leadership has also been shown to play a critical role in encouraging workers to

perform safety behaviours (Hofmann & Morgeson, 1999; Griffin & Talati, 2014; Neal

& Griffin, 2002). Leaders are in the position to motivate their teams to work safely

and help create a safety culture within the organisation by encouraging and rewarding


safety behaviours among employees (Hofmann & Morgeson, 2004; Griffin & Talati,

2014).

Within the field of occupational safety, transactional and transformational

leadership is often used to examine the effect of leadership on safety behaviours

(Clarke, 2013a). Transactional leadership emphasises the exchange process between

leaders and members, where rewards are given to the members when an acceptable

level of effort and performance are displayed, and sanctions are imposed when an

expected level of performance is not met or if undesired behaviours are performed

(Bass, 1997). In contrast, transformational leaders are less focused on the exchange

process with their followers, but instead act as a role model and encourage members

to align their value systems with the overall organisational goals (Bass, 1997).

Previous research has shown that transformational leaders were more effective in

motivating employees to behave safely especially if safety was considered as a core

value promoted by leaders (Barling, Loughlin, & Kelloway, 2002).

More recently, however, researchers have distinguished between the impact of

active transactional leaders and passive leadership (Clarke, 2013a; Jiang & Probst,

2016). A meta-analysis by Clarke (2013) differentiated transactional leadership styles

into active versus passive, arguing that “active leaders monitor subordinate’s

behaviour, anticipate problems and take proactive steps to implement corrective

actions. In contrast, passive leaders wait until the behaviour has created problems

before taking any action” (p. 25). Passive leaders take a more laissez-fare approach,

avoiding responsibilities and making decisions and are usually absent when needed by

subordinates (Jiang & Probst, 2016). Available research on passive leadership showed

that this leadership style is related to poor safety climate and higher frequency of

workplace incidents (Kelloway, Mullen, & Francis, 2006; Zohar, 2002).

Clarke’s (2013) meta-analysis demonstrated that while active transactional

leadership is important in ensuring compliance to safety rules and regulations in

employees, transformational leadership is more essential in encouraging employees to

go beyond the rules and regulations and be proactive about safety issues at work.

However, Clarke’s (2013) review did not study the meta-analytic effect of passive

leadership on safety behaviours. Therefore, the current review built on Clarke’s (2013)

review by examining the impact of passive leadership on proactive safety behaviours,

in addition to transformational and active transactional styles of leadership.


In addition to leadership styles, the role of perceived organisational support and

organisational trust in workplace health and safety are gaining interest in recent years

(e.g., Conchie & Donald, 2009; Conchie, Donald, & Taylor, 2006; Tucker, Chmiel,

Turner, Hershcovis, & Stride, 2008). Perceived organisational support is based on the

organisational support theory, which proposes that, “employees develop global beliefs

concerning the extent to which the organisational values their contributions and cares

about their well-being” (Rhoades and Eisenberger, 2002, p. 698). Thus, perceived

organisational support is the perception that aid will be available from the organisation

when it is needed (Eisenberger, Huntington, Hutchison, & Sowa, 1986; Rhoades &

Eisenberger, 2002). Perceived organisational support had been examined widely

within the organisational psychology literature due to its strong associations with job

satisfaction, wellbeing and job performance (Kurtessis et al., 2017; Rhoades &

Eisenberger, 2002). However, its application within the OHS field has only been

examined sporadically. Initial interest on the role of perceived organisational support

within OHS were introduced by Geller and colleagues in the mid 1990s (Geller et al.,

1996; Roberts & Geller, 1995). They demonstrated that having a caring work

environment promoted safety within the workplace. Geller et al. (2001) found that

employees who felt supported within their organisations were more likely to take

responsibility in ensuring their co-workers safety due to their perceived sense of

camaraderie. From their work, Geller and colleagues argued that actively caring about

employees could act as a supplement or even an alternative approach to monitoring

safety within the workplace (Geller et al., 1996; Roberts & Geller, 1995).

On the other hand, organisational trust describes a more specific interpersonal

process that occurs within the work environment compared to perceived organisational

support. While organisational trust is a complex topic and disagreements exist in

regards to its conceptualisation, it has been defined as a willingness to depend on

another party as well as an expectation of reciprocity (Mayer, Davis, & Schoorman,

1995). The role of trust on safety had been widely recognised by practitioners and

researchers and it is usually seen as an important aspect of safety culture and

transformational leadership (Conchie et al., 2006). For instance, the most widely

accepted and comprehensive definition of safety culture as defined by the Advisory

Committee on the Safety of Nuclear Installations (ACSNI) suggests that mutual trust


is important to develop an effective safety culture10. In addition, Bass’ (1997) model

of transformational leadership posits that this leadership style is effective, in part,

through its facilitation of building trust in others who follow the leader. However,

currently, empirical investigation of trust as its own concept and its influence on OHS

is lacking. Recent works by Conchie and colleagues (Conchie, 2013; Conchie &

Donald, 2009) examined trust towards leaders, workers and organisation, as its own

concept. They found that trust is an important factor in promoting safety behaviour,

especially if such behaviours were not expected in one’s formal role. Trust in

supervisors and co-workers was shown to be important in safety voice and safety

initiatives as workers would be more confident that they would not be “blamed” for

speaking up about safety issues (Conchie et al., 2006). The available evidence thus

suggests that trust is important in fostering open communication within the workplace

and, in particular, with respect to the open reporting of safety incidents (Conchie et al.,

2006).

Safety climate describes the employees’ shared perceptions of safety policies

and procedures and how safety is valued within the organisation. On the other hand,

leadership, perceived organisational support and organisational trust tap into the

impact of interpersonal processes or social exchanges within the organisation and its

potential influence on safety. Social exchange theory is often utilised to explain the

interpersonal processes that occur within the organisation. Social exchange theory

suggests that if one party acts in ways that are beneficial to another party, this creates

an implicit obligation for future reciprocity (Cropanzano & Mitchell, 2005). In

essence, social exchange theory describes the reciprocal relationship between two

parties (Xerri, 2013). Within the workplace, reciprocity refers to the cooperative

exchange between employees or between the organisation and employees (Dabos &

Rousseau, 2004). Social exchange theory has also been previously utilised as the

theoretical framework for addressing safety citizenship behaviours (Hofmann &

Morgeson, 1999; Hofmann, Morgeson, & Gerras, 2003). Hofmann and Morgeson

10 The full definition of safety culture as proposed by the ACSNI is as follows,“…the product of individual and group values, attitudes, perceptions, competencies, and patterns of behaviour that determine commitment to, and the style and proficiency of, an organisation’s health and safety management. Organisations with a positive safety culture are characterised by communications founded on mutual trust, by shared perceptions of the importance of safety and by the efficacy of preventative measures” (Health and Safety Executive, 2005, p. 3).


(1999) argued that if organisations actively demonstrate that they care and value their

employees, then employees would perceive that management is open and accepting of

employees’ safety concerns.

From the organisational climate paradigm, a positive safety climate may

encourage employees to act proactively towards safety due to the shared perception

that safety is valued within the organisation. However, from the social exchange

perspective, proactive safety behaviours could be encouraged via an implied obligation

for workers to act in a safe manner especially if the employees receive support and

trust from their organisations. In other words, if employees feel supported, cared for,

inspired by their leaders and have trust towards their managers, they are more likely

to take initiative in safety matters, especially if their wellbeing and safety is valued

within the organisation.

The current study extends previous reviews on safety behaviours by examining

the role of leadership styles, perceived organisational support and organisational trust

on proactive safety behaviours. Previous meta-analyses have tended to focus only on

the role of safety climate and leadership (Christian et al., 2009; Clarke, 2006, 2013a)

without looking at other organisational factors that may influence employees’ safety

behaviours.


It is only in recent years that interpersonal processes and relationships within

organisations have gained research attention in the work driving safety context (e.g.,

Newnam, Warmerdam, Sheppard, Griffin, & Stevenson, 2017; Peretz & Luria, 2017;

Warmerdam et al., 2018). Available research in work driving safety has identified that

leaders play a critical role in creating a context in which safety is valued and prioritised

within his or her team (Newnam et al., 2012). Although there is very limited research

that has examined the role of leadership styles on work driving safety, available

evidence suggests that having a collaborative relationship between leaders and drivers

is associated with safer driving performance (Newnam et al., 2012). Furthermore,

research on the role of perceived organisational support and organisational trust on

OHS is currently limited, therefore, their role within the specific domain of work

driving safety remains unclear. Peretz and Luria (2017) did not directly examined the

impact of perceived organisational support and trust in work drivers, but they found

that having low quality friendships at work is associated with unsafe driving


behaviours (e.g., sharper turns). Therefore, it is possible that these interpersonal

processes may also influence employees’ engagement on proactive safety behaviours

when driving for work.

Work Demands and Work Autonomy

Work design is also an important aspect of Curcuruto and Griffin’s (2017) model

on safety proactivity. Work design describes “how jobs, tasks, and roles are structured,

enacted, and modified, as well as the impact of these structures, enactments, and

modifications on individual, group, and organizational outcomes” (Grant & Parker,

2009, p. 5). Although the previously discussed antecedents of proactive safety

behaviours examines the role of shared perception of safety and interpersonal

processes within the organisation, work design examines the variables inherent within

the job that may encourage or inhibit one’s engagement in safety behaviours. In other

words, work design could have an impact on whether an employee acts in a proactive

manner towards safety. Studies that examine work design and its influence on safety

performance typically use the job demands and resources model and/or the high

performance systems model11 to explain the relationships between work tasks and

safety behaviour (Parker et al., 2001; Zacharatos, Barling, & Iverson, 2005). These

models usually focus on the role of work demands and work autonomy on safety

behaviours.

Work demands within the OHS literature include risks and hazards, physical

demands and work complexity that employees may experience at work (Nahrgang et

al., 2011). It is possible that the strain and exhaustion that an individual may

experience due to these work demands may reduce their engagement in safety

behaviours (Nahrgang et al., 2011). In Nahrgang et al.’s (2011) meta-analysis of safety

outcomes, they found that high work demands (e.g., risk, hazards, physical demands

and work complexity) explained 20% of the variance in adverse events and unsafe

behaviour. Conversely, having job resources, which may involve work autonomy,

could reduce the associated physiological and psychological strains associated with

work demands (Nahrgang et al., 2011). Having autonomy encourages workers to feel

freedom regarding their work and enables them to achieve their goals (Parker et al.,

11 Parker, Axtell, and Turner (2001) defined high performance work systems as: “autonomous job designs with reasonable role demands supported by adequate training, effective leaders, the flow of information, and employment security are fundamental components” (p. 215).


2001). Autonomy allows employees to feel agency and provide input when necessary.

Indeed, research has found autonomy to be positively related to higher levels of

reporting safety incidents (Parker et al., 2001).

The aforementioned research shows when other demanding tasks require the

focus of workers, and if resources are stretched, safety is often ignored. In these

situations, employees will often narrow their roles to exclude responsibilities outside

their jobs and, therefore, may be less likely to proactively participate in addressing

safety issues at work (Turner, Chmiel, & Walls, 2005; Turner, Stride, Carter,

McCaughey, & Carroll, 2012). Previous meta-analyses have inclined to only examine

the role of perceived work risk on safety compliance (Christian et al., 2009) and the

role of work demands and autonomy on unsafe work outcomes (e.g., unsafe

behaviours, accidents, injuries; Nahrgang et al., 2011). The current meta-analysis

extended the literature by examining the effect of work demands and autonomy on

proactive safety behaviours.


Studies on the role of work demands and work autonomy on proactive safety

behaviours of work drivers are limited. However, available research shows that driving

for work can be highly demanding (e.g., physical workload from extended driving

periods, mental stress from traffic congestion, getting to appointments on time;

Useche, Ortiz, & Cendales, 2017). These work demands have been found to be

associated with aggressive driving when commuting to and from work (Li, Wang, Li,

& Zhou, 2017), poor physical health (e.g., neck and back pains; Krause, Ragland,

Greiner, Syme, & Fisher, 1997) and stress in work drivers (Raggatt, 1991).

While it is clear the driving for work can be highly demanding, employee’s

perceived level of autonomy when driving for work is unclear. Some researchers have

argued that work driving is associated with high levels of autonomy (e.g., work drivers

performing their work tasks at geographically remote locations and outside the

physical constraints of a workplace; Newnam et al., 2017; Zohar, Huang, Lee, &

Robertson, 2015). However, other researchers have suggested that work drivers’

autonomy could be limited by their organisation’s policies and procedures (e.g.,

constraints in job scheduling; Morrow & Crum, 2004). Thus, work demands and work

autonomy could have an impact on work drivers’ safety behaviours but their influence


has yet to be empirically investigated in regards to their impact on one’s engagement

with proactive safety behaviours while driving for work.

4.2.2 Proximal Antecedents

As shown in Figure 4.1, motivational factors and future oriented states are

central to Curcuruto and Griffin’s (2017) model of safety proactivity and act as

proximal antecedents to proactive safety behaviours. Curucuto and Griffin (2017)

argued that several cognitive-mechanisms may impact directly on one’s motivation to

act in a proactive manner. These motivational drivers were referred to as: capability

and commitment drivers. Based on Vroom’s (1964) expectancy-valence theory12,

capability drivers was conceptualised as the ‘can-do’ motivational processes that

explains why individuals may perform proactive safety behaviours (e.g., self-efficacy

and their perceived control). In contrast, commitment drivers tap into individuals’

‘reason-to’ perform these behaviours (e.g., felt responsibility and psychological

ownership of safety within the workplace).

Future oriented states were not conceptualised as motivational factors, but

instead, as an individual’s tendency to look towards future possibilities. In Curcuruto

and Griffin’s (2017) safety proactivity framework, they argued that individuals who

have the propensity to anticipate risks and those who strive to improve safety in the

workplace beyond the minimal standards are more likely to behave proactively for

safety. These individuals have the future orientation of ‘anticipation-orientation’ and

‘improvement-orientation’. Anticipation-oriented and improvement-oriented

individuals are more likely to suggest new or different ways to do things more safely.

These individuals are also more willing to think about ways to improve safety at work

12 In its simplest form, expectancy-valence theory argues that the individual carries out a behaviour depending on (a) the perceived value of the outcome (valence) and (b) the perceived probability of achieving that outcome (expectancy; Pinder, 2008). Curcuruto and Griffin’s (2017) model of safety proactivity argues that the ‘can-do’ motivation to conduct proactive safety behaviours is based on the ‘expectancy’ aspect of Vroom’s expectancy-valence theory, while the ‘reason-to’ motivation to conduct such behaviours taps into the aspect of ‘valence’.

However, the majority of research within the OHS field did not conceptualised safety motivation as a multidimensional construct. It is also recently that the concepts of ‘can-do’ and ‘reason-to’ motivations to act safely have been proposed (Curcuruto et al., 2016; Curcuruto et al. 2017). Therefore, the current review conceptualised safety motivation as a unidimensional construct. Furthermore, concepts that may be associated with safety motivation, such as self-efficacy and perceived control, were conceptualised as its own constructs.


even if activities are running smoothly and there is no evidence of apparent threat

(Curcuruto et al., 2016).

Using Curcuruto and Griffin’s (2017) safety proactivity model as a guide, the

discussions of proximal antecedents within the review examines the role of safety

motivation, regulatory focus, self-efficacy and perceived control on proactive safety

behaviours. Safety knowledge was not originally conceptualised as a proximal factor

of the safety proactivity model, however, previous frameworks (e.g., Griffin and Neal,

2000) considered safety knowledge as a proximal antecedent of safety behaviours.

Therefore, the current review also examined safety knowledge as a proximal

antecedent of proactive safety behaviours.

Safety Knowledge and Safety Motivation

Safety knowledge and safety motivation are usually considered as proximal

antecedents of safety behaviours. Safety knowledge is defined as employees’

understanding of the safety procedures in the workplace (Griffin and Neal, 2000).

Griffin and Neal (2000) argued that individuals must understand how to perform their

work safely and must also have the necessary skills to comply with safety procedures.

Safety motivation, on the other hand, refers to the person’s willingness to put effort in

behaving safely and the value that the person puts on ‘being safe’. Employees may

work in an environment that has a positive safety climate, but they must also be

motivated to participate in safety activities.

The role of safety knowledge and safety motivation on safety behaviours had

been extensively examined in the literature. Both safety knowledge and safety

motivation have been shown to have a direct influence on proactive safety behaviours

(Christian et al., 2009; Curcuruto & Griffin, 2017). In their meta-analysis, Christian et

al. (2009) sought to examine such relationships; however, due to a lack of studies, they

were able to only examine the relationship between safety knowledge and safety

participation. Their review found a large effect between safety knowledge and safety

participation (Mp = .61). The current review extended Christian et al.’s (2009) research

by calculating the meta-analytic effect size between safety motivation and proactive

safety behaviours.


Regulatory Focus, Self-efficacy and Perceived Control

Although the majority of research currently has focused on safety knowledge

and motivation as proximal antecedents of safety behaviours, recent studies have

begun to investigate other possible proximal antecedents, such has employees’

regulatory focus. The self-regulation focus theory describes the two cognitive

approaches that individuals engage in when deciding which behaviours to perform

during a goal pursuit (Higgins, 1997; Henning et al., 2009). These two cognitive

approaches are promotion focused or prevention focused (Higgins, 1997). Promotion

focused individuals seek out ways to accomplish tasks, often engaging in strategies

that maximise rewards (Henning et al., 2009); while, prevention focused individuals

tend to avoid activities that may deter a successful accomplishment of tasks.

Prevention focused individuals mostly engage in strategies that avoid punishment

(Henning et al., 2009) while promotion focused individuals are more likely to take

initiative to change aspects of their immediate work environment (Aryee & Hsiung,

2016). Promotion focus and prevention focus are considered as complementary goal-

oriented states of safety proactivity in Curcuruto and Griffin’s (2017) model in

addition to safety motivation. Curcuruto and Griffin (2017) argued that these

regulatory states enable people to direct attention to future possibilities and the

potential for unexpected events to impinge on current operations adopting more

proactive strategies for a safer environment. At the time of writing, this is the first

study meta-analytic review that examined the role of regulatory focus on proactive

safety behaviours.

Curcuruto and Griffin’s (2017) model also argued that self-efficacy and

perceived control are essential in safety proactivity. Self-efficacy is a person’s level of

confidence in their ability to perform tasks successfully under a range of conditions.

On the other hand, perceived control refers to a person’s belief that their actions

directly influence their immediate surroundings. As mentioned previously, self-

efficacy and perceived control are examples of commitment drivers under Curcuruto

and Griffin’s (2017) model which they describe as the ‘can-do’ motivational processes

that explain why individuals may perform proactive safety behaviours. In other words,

when individuals have the confidence in their ability to be proactive in regards to their

safety and they believe that they have control over their work tasks, they are more

likely to perform proactive safety behaviours. In a study of employees from the


construction and transport industries, Fugas and colleagues (Fugas, Melia, & Silva,

2011; Fugas, Silva, & Melia, 2012, 2013) found that perceived control and self-

efficacy had direct effects on proactive safety behaviours.


Only some of the aforementioned proximal antecedents have been studied within

the work driving context. For instance, while research is limited, studies exist on the

role of safety motivation, safety knowledge, perceived control and self-efficacy on

work driving behaviours. Available studies have suggested that high safety motivation

was associated with lower self-reported crashes (Newnam, Griffin, & Mason, 2008).

Newnam et al. (2008) also found that the safety motivation of drivers was stronger if

the fleet managers and supervisors highly valued driving safety. In the same study,

Newnam et al. (2008) found that self-efficacy was not a significant predictor of work

drivers’ self-reported crashes but it demonstrated a significant association with their

safety motivation. Furthermore, a report by the European Agency for Safety and

Health at Work reviewed several case studies conducted in Europe regarding the

management of risks in road transport drivers (Copsey et al., 2011). The report found

that increasing knowledge of the risks, policies and procedures via training, driver

manuals and information knowledge-sharing, helped to reduce risks when transporting

goods. However, these studies did not examine the influence of safety motivation and

knowledge on work drivers’ proactive safety behaviours.

The evidence regarding the role of perceived control on work driving safety is

mixed. Although Wills et al. (2009) demonstrated that perceived control was not a

significant predictor of unsafe driving behaviours and driving violations in work

drivers (e.g., self-reported crashes and traffic offences), Poulter, Chapman, Bibby,

Clarke, and Crundall (2008) found that perceived control showed a significant direct

effect on work drivers’ adherence with road traffic laws. The difference on the results

between these two studies could be due to the different outcome variables. More

specifically, in Will et al.’s (2009) study, it could be argued that work drivers had less

perceived control of unsafe driving behaviours, especially if their organisation's

climate encourage unsafe practices (i.e., driving while fatigued to meet production

goals). Indeed, in Will et al.’s (2009) study, safety climate showed a stronger effect on

driving behaviours compared to perceived control. Nevertheless, these studies


examined the influence of perceived control on driving behaviours, but not on the

proactive safety behaviours of work drivers.

4.2.3 Summary of the Literature

The previous sections critically synthesised the contextual and proximal

antecedents of proactive safety behaviours and its related constructs, based on

Curcuruto and Griffin’s (2017) framework. It is important to examine these

antecedents as behaviours that go beyond one’s expected roles “do not exist in a

vacuum” (Gyekye & Salminen, 2005, p. 815). The literature specifically focused on

safety climate, organisation-based social exchange and interpersonal processes

(specifically, leadership, perceived organisational support and trust), and work design

(specifically, work demands and autonomy) as contextual antecedents of proactive

safety behaviours. Safety knowledge, safety motivation, regulatory focus (specifically,

promotion focus and prevention focus), self-efficacy and perceived control were

considered as the proximal antecedents. The relevance of these antecedents on work

driving safety were also examined as the review would inform the development of a

research model of the current program of research. Furthermore, since contextual

antecedents were considered to be more distal compared to proximal antecedents, it

was anticipated that proximal antecedents would yield larger relationships with

proactive safety behaviours compared to the distal contextual factors.

4.3 METHOD

4.3.1 Search Strategy

A literature search was conducted to identify empirical studies that had examined

factors that influenced proactive safety behaviours. The first component of the

literature search was the use of electronic databases for any relevant studies that were

written in English and with no restriction imposed on the year of publication. The

databases of PsychINFO (via Ebscohost), Web of Science and Scopus were utilised to

conduct the search. These databases give wide access to literature in behavioural

science (PsychINFO) and cross-disciplinary research (Web of Science and Scopus).

The following keywords for the search were used: “safety citizenship”, “safety

participation”, “extra-role”, “actively caring” and “proactive safety”. The AND search

strategy was used for “extra-role” and “actively caring” with “safety” (e.g., “extra-

role” AND “safety”) to denote that the initial keyword search was specific on safety


issues (see Table 4.1). The AND search strategy was also used for Scopus as the initial

search for the “proactive safety” keyword produced more than 200 results.

Table 4.1 Search Terms used for each database

Database Search Term

PsychINFO (via Ebscohost) “proactive safety AND behavi*”, “safety citizenship”

(abstract), “safety participation”, “actively caring AND

safety”, “extra role AND safety”

Web of Science “proactive safety”, “safety citizenship”, “safety participation”,

“actively caring AND safety”, “extra role AND safety”

Scopus “proactive safety AND behavi*”, “safety citizenship”

(abstract), “safety participation”, “actively caring AND

safety”, “extra role AND safety”

A second component of the literature search was a manual search of journals that

targeted organisational psychology and safety to ensure all relevant articles are

included. The journals that were included in the search were: Journal of Applied

Psychology, Personnel Psychology, Journal of Organizational Behavior, Journal of

Occupational Health Psychology, Safety Science, Journal of Safety Research and

Accident Analysis and Prevention. Furthermore, the reference lists of the retrieved

articles and previous meta-analyses on safety behaviours and safety performance (e.g.,

Christian et al., 2009; Clarke, 2006, 2013) were reviewed for additional studies. Figure

4.2 shows the search strategy that was utilised for this study.


4.3.2 Inclusion and Exclusion Criteria

To be included in the review, studies were required to meet three criteria:

1. Studies had to have included a measure of proactive safety behaviour (Table

4.2). Studies were excluded if they only used an overall measure of safety

performance without proving a distinction between proactive safety

behaviour and safety compliance.

Papers excluded: Duplicates n = 246

Papers for full-text review n = 130

Studies included in the data extraction

n = 78

Papers excluded: Did not explicitly study proactive behaviours n = 37 Did not provide correlations or statistics to calculate an effect size n = 9

Cannot access full-text n = 6

Papers for title and abstract review

n = 253

Manual search n = 52 Database search n = 447

Papers excluded: Inclusion criteria not met n = 123

Total papers n = 499

Figure 4.2 Search strategy for the systematic review and meta-analysis

Studies included in the meta-analysis calculation

n = 71

Papers excluded in the meta-analytic coding n = 7


Table 4.2 Terminologies and measures used for the current study

Terminology / Construct Definition Behaviours measured Safety Initiative (Andressien, 1997; Simard & Marchard, 1995)

Employees’ attempts to create a safer work environment.

Employees take initiatives in order to work more safely Make suggestions in order to improve safety at work environment Put pressure on the supervisor to improve safety at work

Actively caring about safety (Geller, Johnson, Redmon, & Mawhinney, 2001; Geller, Roberts, & Gilmore, 1996)

Employees acting to benefit the safety of other employees.

Observing and recording the safe and unsafe behaviours of coworkers, and then provide constructive behavioral feedback

Willingness to participate in health and management programs (Cree & Kelloway, 1997)

By participating in health and safety programs, employees have the opportunity to have a "voice" in the management of workplace health and safety.

Participation in various programs (e.g., attending safety meetings, attending safety training).

Safety Participation (Griffin & Neal, 2000)

Safety behaviours that may not necessarily directly contribute to workplace safety, but help develop an environment that supports safety.

Promoting safety program Extra effort to improve safety Voluntarily carry out tasks to improve safety

Safety Citizenship (Hofmann et al., 2003)

Safety behaviours that are not formally rewarded by the organisation, but plays a strong role in accident prevention due to its facilitation in improving the safety and learning within the workplace.

Whistle-blowing Safety voice Helping Safety initiative

Proactive safety behaviours (Curcuruto et al., 2015)

Safety behaviours that are challenging in nature and seek to bring out positive change in the safety practices within the organisation.

Whistle-blowing Safety voice Safety initiative Keeping informed

Prosocial safety behaviours (Curcuruto et al., 2015)

Safety behaviours that are affiliative in nature. This behaviour generally manifest itself as helping co-workers and looking out for their welfare and safety. Often seen as a way to foster good relationships.

Helping Stewardship

2. Studies had to have been conducted within an organisational setting. Thus,

studies that utilised a student population as the study sample were excluded.

3. Studies were required to have reported sample sizes and correlations (or

statistics that could be transformed into correlations) between proactive

safety behaviour and an antecedent variable. Qualitative studies and case

studies were excluded in the meta-analysis.


After the data extraction, some studies were not added to the calculations of the

meta-analytic correlations due to the small number of studies examining the same

antecedent. For instance, the study by Zhang and Wu (2014) on mindfulness and

general mental ability were excluded in the meta-analytic calculations. Furthermore,

some studies examined proactive safety behaviours and related constructs without a

discussion of any antecedents. From this process, a total of 7 studies that were

extracted for full-text review were not included in the meta-analytic calculations.

Therefore, a total of 71 studies were included for the meta-analytic calculations.

4.3.3 Coding

Sample size, the country where the study was performed, industry from which

drivers were from, study design, measure and definition of proactive safety behaviour,

and the antecedents were coded. The coding of the antecedents is available in

Appendix A. Effect sizes were recorded for all possible statistical relations among the

antecedents and proactive safety behaviour variables in order to calculate the pooled

estimate effect size. Accurate coding is essential for meta-analysis. Once the coding

was finished, 15% of the sample were randomly selected, cross-checked with the

original sources for inaccuracies by the same author (the PhD candidate). Inaccurate

input were subsequently revised.

4.3.4 Meta-analysis calculations

Meta-analytic calculations were carried out using Hunter and Schmidt’s random-

effects model (Schmidt & Hunter, 2014). This model was applied due to the different

measurements used for the antecedent and the outcome variable (e.g., safety

participation measure, safety citizenship measures, actively caring measures, see Table

4.2 for the measures coded within the meta-analysis). Data was analysed using the

metafor package in R, with the H&S-estimator (Viechtbauer, 2010). Hunter and

Schmidt’s approach is an appropriate method when estimating pooled effect sizes on

studies that use continuous measures (e.g., correlation coefficient or standardized

mean differences; Borenstein, Hedges, Higgins, & Rothstein, 2009). This approach is

also often used when conducting meta-analysis in the field of the organisational

sciences (Kepes et al., 2013).

Using Hunter and Schmidt’s meta-analytic approach, the effect size between

variables were represented by the observed correlation coefficients (i.e., Pearson’s r)


and weighted by the sample size (Borenstein et al., 2009). These effect sizes were then

corrected with the study’s imperfections or ‘artefacts’ to provide a better estimate of

the effect (Borenstein et al., 2009). Measurement errors tend to attenuate (i.e., lower

or dilute) the observed correlation coefficients.

When using Hunter and Schmidt’s (2004) method, the observed effect sizes are

usually corrected using the measurement errors for the antecedent and outcome

variables by giving more weight when calculating the pooled effect size estimates.

Since the majority of the studies reported the measures’ Cronbach α, these reliability

estimates were used for the corrections using equation 4.1. For studies that did not

report the Cronbach alpha’s for the measures used, the average of the reported

reliability estimates were utilised, an approach recommended by Hunter and Schmidt

(2004).

∙ ∙ Equation

4.1

rcor = corrected r; robv = observed r; rxx = Cronbach α of antecedent variable; ryy = Cronbach α of outcome variable

Other corrections were also applied. Within the dataset, there were studies that

reported multiple estimates of the same relationship. For instance, estimates for several

dimensions of safety climate (e.g., management commitment, safety over production)

were reported instead of providing the overall correlation estimate between safety

climate and proactive safety behaviour13. For these cases, the observed effect sizes

were combined into a single correlation using the average mean of the reported

estimates. This approach prevented a study from being double-counted in the meta-

analytic calculations. While using the composite formula is the preferred method for

calculating a single correlation for interdependent effect sizes (Geyskens et al., 2009),

some calculations for the composite scores were not possible14. For consistency in

13 For example, a study may provide an effect size for management commitment and safety participation, r = .30 and another effect size for safety over production r = 40, which are dimensions of safety climate. If the study did not provide an overall effect size for safety climate, the average of the available effect sizes that measure the same construct (which is safety climate for this example) was utilised (e.g., average of r = .30 and r = .40 is .35). 14 According to Geyskens, Krishnan, Steenkamp, and Cunha (2009), calculating a single composite score formula is the preferred approach, which follows the formula: composite score = Σrxy/SQRT(k + k*(k -1)*(ΣrY1Y2/k))


reporting, average of estimates was calculated. In contrast, studies that included

multiple independent samples were separately coded.

The following information was collected for the meta-analytic calculations: the

observed correlations, robv, the sample size (N), and the Cronbach’s α of the

antecedents (rxx) and outcome variables (ryy) to correct for the observed Pearson’s r

correlations. Once all information were collected and coded in Microsoft Excel, the

metafor package in R was used to calculate the corrected correlations, rcor, the pooled

observed and corrected correlations estimates and the corresponding standard errors

for the pooled corrected estimates (SE) and the 95% confidence intervals (CI). A 95%

CI excluding zero indicates that one can be 95% confident that the average true

correlation is larger than zero. Therefore, an exclusion of zero within the confidence

intervals suggest a significant result. Confidence intervals provide an estimate of the

variability around the estimated pooled corrected meta-analytic correlations. The total

sample size and the number of studies (k) included in the pooled correlations were

presented. Finally, the heterogeneity of the studies were assessed.

4.4 RESULTS

4.4.1 Study Characteristics

The studies that were analysed for the current meta-analysis are presented in

Table 4.3. The studies were conducted in various industries: agriculture, construction,

healthcare, manufacturing, mining, transport, and the utility, petroleum and energy

sector. The most common industry where the studies were conducted was the utility,

petroleum and energy sector (n = 16; 22.5%). Ten studies used employees from various

industries, retail, military or firefighting (14.1%). Around two-thirds of these studies

were conducted in Europe (n = 22; 31%) and Northern America (n = 25; n = 35%).

The majority of the studies were cross-sectional (n = 65; 95%) and two of these

studies were carried out in two waves. Only 3 studies used a longitudinal design. Just

over half of the studies used the Neal and Griffin’s (2000) safety participation (n = 36;

51%) and 15 (21%) of the studies used Hofmann et al.’s (2003) safety citizenship

where k = no. of correlations and rY1Y2 are the inter-correlations of the composite. While this method is the preferred approach (Geyskens et al., 2009), some data required for the formula were still missing, therefore, available effect sizes were averaged so as to be dealt with in a consistent manner.


behaviours to measure proactive safety behaviours. The remainder of the studies either

developed their own measures (n = 6; 8%) or relied upon a mixture of items adapted

from safety participation and safety citizenship measures (n = 12; 17%).

4.4.2 Meta-analytic Correlations

Table 4.4 demonstrates the pooled meta-analytic correlations between the

perceived work environment, individual differences and the proximal antecedents of

proactive safety behaviours. The Cochran’s Q-Statistics demonstrate heterogeneity in

all extracted relationships between the antecedents and proactive safety behaviours

except for group-level safety climate and passive leadership. Its corresponding I2

denotes the extent of the heterogeneity within the dataset. These results suggest that

the use of a random-effects estimation was appropriate for the calculations of the

pooled estimate effect size. The Egger’s regression tests were conducted to provide an

objective measure for potential publication bias. Egger’s test was used due to its better

suitability for smaller meta-analysis sample sizes (k < 25; Quintana, 2015). From this

test, only the effect size for perceived control showed apparent publication bias (i.e.,

stronger effect sizes being published; Quintana, 2015) due to the significant Egger’s

test.


Table 4.3 Studies included in the data-analysis

Study No.

Author (Year) Industry Country Area Sample Characteristics

Study Design Terminology Used Measure Used Predictors

1 Agnew, Flin, and Mearns (2013)

Healthcare UK Europe 1866 clinical staff Cross-sectional

Safety participation Neal and Griffin's Safety Participation

Safety Climate

2 Andriessen (1978) Construction Netherlands Europe 207 construction workers

Cross-sectional

Safety initiative Original Safety Motivation

3 Aryee and Hsiung (2016)

Others Taiwan Asia 235 firefighters; 91 supervisors

Longitudinal (6m duration)

Safety initiative Hofmann et al.'s Safety Citizenship

Regulatory Focus

4 Barbaranelli, Petitta, and Probst (2015)

Others US and Italy Mixed Various workers Cross-sectional


Safety Climate, Safety Knowledge, Safety Motivation

5 Boughaba, Hassane, and Roukia (2014)

Utility, Petroleum and Energy

Algeria Middle East

300 workers from an oil company (Company A); 208 workers from an oil company (Company B)

Cross-sectional


Safety Climate

6 Brondino, Silva, and Pasini (2012)

Manufacturing Italy Europe 991 blue collar workers from a metal and mechanical sector

Cross-sectional


Safety Climate

7 Burke, Sarpy, Tesluk, and Smith-Crowe (2002)


US North America

132 hazardous waste workers

Cross-sectional

Engaging in work practices to reduce risk

Original Safety Knowledge

8 Burt, Sepie, and McFadden (2008)

Construction New Zealand Oceana 80 workers from forestry and construction industries

Cross-sectional

Care scale Geller's Actively Caring

Perceived Organisational Support

9 Chen and Chen (2014)

Transport Taiwan Asia 339 flight attendants Cross-sectional

Extra-role safety behaviour; Upward communication

Others or Mixture Job Autonomy

10 Chughtai (2015) Healthcare Pakistan Asia 179 doctors Cross-sectional


Leadership; Self-efficacy

11 Clarke and Ward (2006)

Manufacturing UK Europe

83 team members; 22 supervisors from a glassware manufacturing company

Cross-sectional


Safety Climate, Leadership, Job Autonomy

12 Conchie and Donald (2009)

Construction UK Europe 139 workers and 33 supervisors from a construction company

Cross-sectional

Safety citizenship behaviours

Hofmann et al.'s Safety Citizenship

Leadership, Trust

13 Conchie, Taylor, and Donald (2012)


UK Europe 150 employees; 79 supervisors from an oil company

Cross-sectional

Safety voice Hofmann et al.'s Safety Citizenship

Leadership, Trust, Safety Motivation


Study No.



14 Conchie (2013) Construction UK Europe 251 employees from a construction company

Cross-sectional

Challenge and Affiliative


Leadership, Trust, Safety Motivation, Regulatory Focus

15 Cree and Kelloway (1997)

Manufacturing Canada North America

130 production employees

Cross-sectional

Willingness to participate

Others or Mixture Work demands

16 Cullen and Hammer (2007)

Healthcare US North America

243 nurses and health care workers

Cross-sectional


Work demands

17 Curcuruto, Mearns, and Mariani (2016)

Manufacturing Italy Europe 523 chemical and manufacturing

Exploratory and Validation

Proactive safety Hofmann et al.'s Safety Citizenship

Leadership

18 Eklöf (2002) Agriculture Sweden Europe 92 fishermen Pilot of survey

Activity in safety work Others or Mixture Safety Knowledge; Perceived Control

19 Ford and Tetrick (2011)

Healthcare US North America

171 hospital workers Cross-sectional



Safety Climate, Perceived Control

20 Fugas, Melia, Silvia (2011)

Transport Italy Europe

129 operational workers in a transportation company

T1 and T2 cross-sectional



Safety Climate

21 Fugas, Silva, and Melia (2012)

Construction Italy Europe

356 operational workers in a transportation company

Cross-sectional



Safety Climate, Perceived Control

22 Geller, Roberts, and Gilmore (1996)

Manufacturing US North America

374 workers from a plastic manufacturing plant; 218 workers from a textile manufacturing plant

T1 and T2 cross-sectional

Willingness to actively care

Geller's Actively Caring

Perceived Organisational Support, Perceived Control, Work Demands

23 Griffin and Neal (2000)

Manufacturing AUS Oceana 1264 workers from a construction and mining company

Cross-sectional



24 Griffin and Hu (2013)

Others AUS Oceana 267 workers from various industries

Cross-sectional


Leadership

25 Guo, Yiu, and Gonzalez (2016)

Construction NZ Oceana 213 construction workers

Cross-sectional

Safety participation Original Safety Climate, Safety Knowledge, Safety Motivation

26 Hoffmeister et al. (2014)


US North America

1167 pipefitters and plumbers

Cross-sectional


Safety Climate, Leadership

27 Hofmann and Adam (1996)


US North America

222 workers from a chemical processing plant

Cross-sectional

Approach intention Original Perceived Organisational Support


Study No.



28 Hofmann, Morgeson, and Gerras (2003)

Others US North America

101 workers from a military unit

Cross-sectional




29 Hon, Chan, and Yam (2014)


HK Asia

396 repair, maintenance, minor alternation and addition employees

Cross-sectional


Safety Climate

30 Hsu, Lee, Wu, and Takano (2008)

Utility, Petroleum and Energy Taiwan and

Japan Asia

295 oil refinery workers from Taiwan; 256 oil refinery workers from Japan

Cross-sectional

Willingness to report Original

Safety Climate, Perceived Organisational Support, Work Demands, Regulatory Focus; Self-efficacy

31 Inness, Turner, Barling, and Stride (2010)


159 moonlighters Cross-sectional


Leadership, Work Demands

32 Jiang, Yu, Li, and Li (2010)


China Asia 631 workers from petroleum and chemical industries

Cross-sectional


Safety Climate

33 Jiang and Probst (2016)

Transport US North America

389 workers from a public transit agency

Cross-sectional


Leadership, Safety Knowledge, Safety Motivation

34 Kark, Katz-Navon, and Delegach (2015)

Others Israel Middle East

99 employees from an online survey firm with access to workers; 789 workers and 49 managers in teams of technician and mechanics

Cross-sectional

Safety initiative Others or Mixture Leadership

35 Kath, Marks, Raney (2010)


548 railway workers Cross-sectional

Upward safety communication


Safety Climate, Leadership, Perceived Organisational Support

36 Keffane (2015) Others France Europe 165 workers; 135 managers

Cross-sectional



37 Krauss and Casey (2014)


Australia Oceana 254 workers from a utilities company

Cross-sectional




38 Lievens and Vlerick (2014)

Healthcare Belgium Europe 152 nurses Cross-sectional


Leadership, Safety Knowledge

39 Lu and Yang (2010) Transport Taiwan Asia 336 passenger ferry operators

Cross-sectional


Leadership


Study No.



40 Lu and Yang (2011) Transport Taiwan Asia 155 passenger ferry operators

Cross-sectional


Safety Climate

41

Martinez-Corcoles, Schobel, Gracia, Tomas, and Peiro (2012)

Utility, Petroleum and Energy Spain Europe

495 workers from a nuclear power plant

Cross-sectional


Leadership

42

Martinez-Corcoles, Gracia, Tomas, Peiro, and Schobel (2013)


Spain Europe 479 workers from a nuclear power plant

Cross-sectional


Leadership

43 McGonagle, Childress, Walsh, and Bauerle (2016)

Construction USA North America

290 workers in physical labour jobs - taken with some university students but also field recruitment

Cross-sectional

Safety participation Others or Mixture Safety Climate, Safety Motivation

44 Mearns and Reader (2008)


UK Europe 703 workers from an installation company

Cross-sectional

Safety citizenship Others or Mixture Perceived Organisational Support

45 Mullen (2005) Manufacturing Canada North America

103 workers from manufacturing company; 75 students from service industries

Cross-sectional

Willingness to participate

Others or Mixture Perceived Organisational Support

46 Mullen and Kelloway (2009)

Healthcare Canada North America

54 leaders; 115 healthcare workers

Longitudinal; Pre-post


Self-efficacy; Perceived Organisational Support

47 Mullen, Kelloway, and Teed (2011)

Healthcare Canada North America

241 workers taken from a university sample; 491 healthcare employees from Sample B

Cross-sectional



48 Neal, Griffin, and Hart (2000)

Healthcare Australia Oceana 525 healthcare workers

Cross-sectional



49 Neal and Griffin (2006)

Healthcare Australia Oceana

430 healthcare workers in T1; 490 healthcare workers in T2; 301 healthcare workers in T3

Longitudinal Safety participation Neal and Griffin's Safety Participation

Safety Climate, Safety Motivation

50 Paolillo, Silva, and Pasini (2016)

Manufacturing Italy Europe 481 workers from a metal-mechanical industry

Cross-sectional


Safety Motivation


Study No.



51 Parboteeah and Kapp (2008)


237 workers from an automotive manufacturing company

Cross-sectional


Safety Motivation

52 Probst, Graso, Estrada, and Greer (2013)


128 workers from pulp and paper mill company

Two-Wave Safety citizenship behaviours

Neal and Griffin's Safety Participation

Safety Motivation, Regulatory Focus

53 Sampson, DeArmond, and Chen (2014)


US North America

120 pipefitters belonging in a union local

Cross-sectional

Safety participation Hofmann et al.'s Safety Citizenship

Perceived Organisational Support, Job Autonomy

54 Shea, De Cieri, Donohue, Cooper, and Sheehan (2016)

Others Australia Oceana 3605 workers from various organisations

Cross-sectional


Safety Motivation; Perceived Control

55 Simard and Marchand (1994)


2169 various workers from manufacturing plants

Cross-sectional

Safety initiatives Others or Mixture Safety Climate

56 Simard and Marchand (1995)


23615 workers from manufacturing plants

Cross-sectional

Safety initiatives Others or Mixture

Safety Climate, Perceived Organisational Support, Job Demands

57 Sinclair, Martin, and Sears (2010)


535 retail employees Cross-sectional


Safety Climate, Work Demands, Safety Knowledge, Safety Motivation

58 Smith and DeJoy (2014)


398 full time professional fire-fighters

Cross-sectional


Safety Climate

59 Smith, Eldridge, and DeJoy (2016)


398 full time professional fire-fighters

Cross-sectional



60

Subramaniam, Mohd Shamsudin, Mohd Zin, Sri Ramalu, and Hassan (2016)

Manufacturing Malaysia Asia 120 workers from manufacturing industries

Cross-sectional

Safety participation Others or Mixture Safety Climate

61 Tharaldsen, Mearns, and Knudsen (2010)


UK and Norway

Europe

170 workers from a drilling company in the UK; 621 workers from a drilling company in Norway

Cross-sectional


Trust


Study No.



62 Toderi, Gaggia, Mariani, Mancini, and Broccoli (2015)

Construction Italy Europe 277 workers in construction industry

Cross-sectional


Safety Knowledge, Safety Motivation

63 Tucker, Chmiel, Turner, Hershcovis, and Stride (2008)

Transport UK Europe 213 bus drivers Cross-sectional

Employee safety voice Hofmann et al.'s Safety Citizenship

Perceived Organisational Support

64 Turner, Chmiel, and Walls (2005)

Transport UK Europe 334 railway workers Cross-sectional



Perceived Organisational Support, Job Autonomy

65 Turner, Stride, Carter, McCaughey, and Carroll (2012)

Healthcare UK Europe 280 healthcare workers

Cross-sectional


Job Autonomy

66 Vinodkumar and Bhasi (2010)

Agriculture India Asia 1566 fertilizer companies

Cross-sectional

Safety participation Original Safety Climate, Safety Knowledge, Safety Motivation

67 Willis, Brown, and Prussia (2012)


US North America

821 employees from an electric utility company

Cross-sectional


Others or Mixture Safety Climate

68 Xuesheng and Xintao (2011)

Mining China Asia 450 workers from a coal mine

Cross-sectional



Safety Climate

69 Yuan, Li, and Lin (2014)


Cross-sectional


Safety Climate

70 Yuan, Li, and Tetrick (2015)


Cross-sectional


Perceived Organisational Support; Self-efficacy

71 Zacharatos, Barling, and Iverson (2005)


Canada North America

191 workers from telecoms and petroleum companies

Cross-sectional

Safety initiative Others or Mixture Safety Climate, Trust, Safety Knowledge, Safety Motivation


Table 4.4 Results of the meta-analysis

Construct N k robv rcor SE (rcor) 95%

CIlower 95%

CIupper Q I2

Egger’s Test

Contextual Antecedents Safety climate 20,886 47 0.35 0.43 0.03 0.37 0.49 454.67** 89.5% p = 0.913

Psychological safety climate 17,368 38 0.35 0.43 0.04 0.36 0.51 428.07** 90.9% p = 0.895 Group-level safety climate 2,170 6 0.39 0.47 0.03 0.42 0.52 7.64 16.0% p = 0.990 Organisational-level safety climate 1,348 3 0.30 0.37 0.07 0.23 0.50 10.59** 64.3% p = 0.85

Leadership Transformational leadership 8,518 28 0.33 0.37 0.04 0.30 0.44 163.04** 82.3% p = 0.936 Transactional leadership 3,534 9 0.18 0.22 0.08 0.06 0.38 64.37** 83.3% p = 0.932 Passive leadership 1,011 3 -0.18 -0.23 0.04 -0.31 -0.16 2.37 0.0% p = 0.136

Perceived organisational support 4,865 15 0.27 0.33 0.04 0.25 0.41 52.29** 70.2% p = 0.924 Organisational trust 1,518 9 0.28 0.32 0.06 0.21 0.43 42.08** 78.3% p = 0.063 Work autonomy 1,115 4 0.38 0.44 0.13 0.19 0.69 67.26** 93.9% p = 0.815 Work demands 3,315 12 0.01 -0.00 0.07 -0.14 0.14 80.28** 84.1% p = 0.872

Proximal Antecedents Perceived knowledge and skill 7,498 16 0.51 0.64 0.04 0.57 0.71 106.95** 84.2% p = 0.147 Safety motivation 11,764 21 0.44 0.53 0.05 0.42 0.63 265.04** 91.2% p = 0.460 Self-efficacy 1,179 5 0.33 0.42 0.53 0.32 0.53 12.84* 61.0% p = 0.475 Perceived control 4,754 6 0.60 0.75 0.12 0.50 0.99 188.02** 93.1% p < .001 Regulatory focus

Promotion focus 2,37

3 9 0.32 0.38 0.07 0.24 0.51 50.43** 80.5% p = 0.512

Prevention focus 1,482 4 0.03 0.03 0.07 -0.12 0.18 20.45** 72.2% p = 0.664 Notes. k = number of studies; robv = Mean pooled observed correlation coefficients; rcor = Mean pooled corrected correlation coefficients; SE

(rcor) = Standard Error of the Mean pooled corrected correlation coefficients; CI = Confidence Intervals; Q = Cochran’s Q-Statistics; I2 = I2 statistics.



Given that the studies included in the analysis generally featured safety climate

and leadership when examining safety behaviours, it is not surprising that these two

factors were the two most studied antecedents of proactive safety behaviour within the

current review. Safety climate was the most studied antecedent of proactive safety

behaviour with individually coded effect sizes of k = 47 and while there were k = 40

individually coded effect sizes for leadership styles.

In line with previous reviews (Christian et al., 2009; Clarke, 2006), the current

findings showed a moderate and positive relationship between safety climate and

proactive safety behaviours (rcor = 0.43; 95% CI [0.37, 0.49]) and this effect was

significant. When the different levels of safety climate were examined, all levels

showed significant and moderate relationships with proactive safety behaviours,

however, group-level safety climate showed the strongest effect (rcor = 0.47; 95% CI

[0.42, 0.52]).

The current study also examined the effects of three different leadership styles

(transformational, transactional and passive) on proactive safety behaviours.

Transformational leadership was positively related to proactive safety behaviour (rcor

= 0.37; 95% CI [0.30, 0.44]), while transactional was significantly related to proactive

safety behaviours (rcor = 0.22; 95% CI [0.06, 0.38]) but its effect was weaker compared

to transformational leadership. The analysis revealed that passive leadership was

significantly and negatively related to proactive safety behaviours (rcor = -0.23; 95%

CI [-0.31, -0.16]).

Overall, both perceived organisational support (rcor = 0.33; 95% CI [0.25, 0.41])

and trust (rcor = 0.32; 95% CI [0.21, 0.43]) were positively related to proactive safety

behaviours, with similar effect sizes. When aspects of work design were analysed,

work autonomy showed a moderate and positive relationship with proactive safety

behaviours (rcor = 0.44; 95% CI [0.19, 0.69]). The pooled estimate effect was

significant. On the other hand, work demands showed a very low and negative

relationship with proactive safety behaviours (rcor = -0.00; 95% CI [-0.06, 0.14]), and

the effect was not significant.


4.4.4 Proximal Antecedent

The current meta-analysis revealed positive effect sizes among the extracted

proximal antecedents and proactive safety behaviours. Most effects were significant,

except for prevention focus. Perceived control showed the strongest effect (rcor = 0.75;

95% CI [0.50, 0.99]). However, the strength of the pooled estimate effect size for

perceived control should be interpreted with caution due to the significant Egger’s test.

Safety motivation also showed a strong effect with proactive safety behaviours (rcor =

0.53; 95% CI [0.42, 0.63]) as well as perceived knowledge and skills (rcor = 0.64; 95%

CI [0.57, 0.71]). Self-efficacy showed a moderate to strong effect size (rcor = 0.42;

95% CI [0.32, 0.53]).

When the differential effect of regulatory focus was analysed, the results

revealed that promotion focus showed a moderate and significant pooled estimate

effect (rcor = 0.38; 95% CI [0.24, 0.51]) while prevention focus showed a very weak

and non-significant effect (rcor = 0.03; 95% CI [-0.12, 0.18]).

4.4.5 Selected Studies Relevant to Work Driving Safety

Out of the 71 empirical studies that were extracted and analysed in the current

meta-analysis, one study explored workplace safety in terms of participating in road

safety practices (Keffane, 2015) and another study examined safety voice using a

sample of urban bus drivers (Tucker et al., 2008). Keffane (2015) examined the role

of safety climate in management of road safety practices using a sample of employees

who drive for work in France. Keffane’s study (2015) utilised established measures

from safety literature which were adapted to suit the work driving context. The path

analysis revealed that safety communication and feedback was the only dimension of

safety climate that demonstrated a significant positive effect on participation in road

safety practices. Other dimensions that were examined (e.g., safety commitment,

safety training) were not significant predictors of safety participation.

While Tucker et al.’s (2008) study did not specifically examine the context of

work driving safety, they looked at safety voice behaviours in bus drivers. They found

that perceived co-worker and organisational support for safety were significant

predictors of employee voice. Bus drivers spoke out more about safety issues when

they perceived that their organisation supported safety and this relationship was

mediated by employees’ perceptions of their co-workers’ supporting workplace safety.


Co-workers support for workplace safety also showed a stronger effect on drivers’

safety voice compared to organisation-level support for safety.

4.5 DISCUSSION

Utilising Curcuruto and Griffin’s (2017) model of safety proactivity, the current

study undertook a meta-analytic investigation of factors that facilitate and inhibit

proactive safety behaviours. Specifically, it examined the contextual antecedents (i.e.,

safety climate, leadership, perceived organisational support, trust, work autonomy and

work demands) and proximal antecedents (safety knowledge, safety motivation, self-

efficacy, perceived control and regulatory focus) of proactive safety behaviours. The

purpose of the current review was to inform the development of a research model of

proactive safety behaviour that would be applied within the work driving context

(which is detailed in Chapter 6).

The present study has a number of theoretical implications. First, the current

meta-analysis further confirms the relationship between safety climate and proactive

safety behaviour. This finding is in line with previous reviews that have also

demonstrated that a positive safety climate is associated with higher levels of active

participation in workplace safety practices (Christian et al., 2009; Clarke, 2006).

Employees’ perceptions of their workplace safety climate influence their behaviour.

Therefore, in a positive safety climate, employees are encouraged to proactively

participate in safety activities and are likely to suggest safety initiatives within the

workplace.

Didla, Mearn and Flin (2007) argued that when a work environment place

emphasis on safety over production and the importance of safety values are

incorporated in all levels of the organisation (e.g., top-level, sub-units), employees are

encouraged to take more initiatives in ensuring workplace safety. The current study

extended the literature by segregating the different levels of safety climate (i.e.,

psychological-level, group-level or organisation-level) and their influence on

proactive safety behaviours. The current findings clearly demonstrated that, when the

shared perception of safety is analysed within the work-group or direct supervisor level

(i.e., group-level), the effect of safety climate on proactive safety behaviours were

stronger compared to shared perception of safety among the top organisational level.

The current meta-analysis provides further evidence to the multi-level nature of safety


climate (Zohar, 2008; Zohar & Luria, 2005) and that supervisors’ safety practices may

have a stronger impact on employees’ initiatives to promote safety at work.

In regards to the effect of leadership styles, the current meta-analysis revealed

that transformational leadership has a moderate and positive relationship with

proactive safety behaviours while transactional leadership showed a positive but

weaker effect. On the other hand, passive leadership demonstrated a negative effect on

proactive safety behaviours. These findings provide further support that leadership is

an important antecedent to proactive safety behaviours. More specifically,

transformational leadership style has the strongest effect in ensuring proactive safety

behaviours in the workplace out of the three leadership styles. This finding is inline

with previous reviews (Clarke, 2006, 2013a).

The current meta-analysis also provided further support on the distinctiveness

between the three types of leadership and the differential effects that they have on

proactive safety behaviours. It can be argued that although transactional leadership

style was associated with a smaller effect on proactive safety behaviours compared to

transformational leadership, its significant positive effect suggests that transactional

leadership is still a critical factor in safety promotion initiatives in the workplace –

providing that the leader is active (e.g., one who monitors and provides feedback on

errors). The current meta-analysis extends previous reviews on leadership and safety

behaviours, by demonstrating that passive leadership showed a negative effect on

proactive safety behaviours. It is possible that previous studies that found a weak effect

between transactional leadership and safety behaviours (e.g., Kark, Katz-Navon, &

Delegach, 2015) did not differentiate active transactional leaders from passive leaders.

The current findings further clarified that active transactional leaders still play a

positive role in employees’ proactive safety behaviours. However, the passive

leadership style may actually have detrimental effects on workers’ safety behaviours,

with employees having diminished proactivity towards safety.

Furthermore, the current meta-analysis demonstrated that perceived

organisational support, organisational trust and work autonomy have moderate and

positive effects on proactive safety behaviours. While there is currently limited

research on these concepts and their impact on occupational safety, emerging evidence

suggests that individuals must feel supported within their workplace, must trust their

organisations, and must feel autonomous regarding their work in order to be proactive


towards occupational safety (e.g., Mearns et al., 2008; Parker et al., 2001; Turner et

al., 2005; Zacharatos et al., 2005). Having a supportive environment could create a

sense of group cohesion within the workplace and contribute to employees ‘feeling

safe’ to take safety initiatives at work (Geller et al., 1996). Furthermore, being

supported at work could encourage employees to care about the wellbeing and safety

of their co-workers and may encourage their engagement in safety behaviours that

ensure the safety of their co-workers, even when it is not expected (Burt, Sepie, &

McFadden, 2008; Geller et al., 1996). Organisational trust, on the other hand, could

foster open communication and enhanced cooperation, which could then encourage

employees to engage in proactive safety behaviours (Conchie & Donald, 2008;

Conchie et al., 2006).

The results also revealed that work autonomy is a critical factor for employees’

proactive safety behaviours. Among the contextual antecedents, work autonomy had

the highest meta-analytic effect on proactive safety behaviours. Curcuruto and Griffin

(2017) argued that employees with higher autonomy may present more opportunities

and motivation to get involved in safety tasks that fall outside their job description.

This could explain the high meta-analytic effect found in the current study. On the

other hand, work demands showed a non-significant effect on proactive safety

behaviours. It is possible that the finding is due to the majority of the studies showing

very low correlations between work demands and proactive safety behaviours.

Employees may feel overwhelmed by the demands they experienced at work and, as

such, they are not motivated or have less energy, to take safety initiatives at work

(Sampson, DeArmond, & Chen, 2014).

Within the proximal antecedents, the current meta-analysis provided further

evidence that safety knowledge and safety motivation are critical proximal antecedents

to proactive safety behaviours. However, the study has also extended current

knowledge by demonstrating that self-efficacy, perceived control and promotion focus

can also act as additional proximal antecedents to proactive safety behaviours. The

previous meta-analysis by Christian et al. (2009) showed that employees need to be

motivated and have the knowledge to engage in proactive behaviours but the current

review extended the literature by demonstrating that having high self-efficacy and

perceived control and being promotion focused also encourage individuals to engage

in such behaviours. Perceived control showed the strongest pooled correlation estimate


on proactive safety behaviours, however, due to the significant publication bias test,

this meta-analytic strength should be interpreted with caution.

The findings of the current study also extended Curcuruto and Griffin’s (2017)

safety proactivity model by providing meta-analytic evidence on their proposed

antecedents of proactive safety behaviours. In addition, the effect sizes of the proximal

antecedents generally showed higher pooled correlation strengths with proactive safety

behaviours compared to the contextual antecedents. This observation gave further

evidence to the suitability of the variables’ groupings within the distal and proximal

categories and the relative ‘distance’ of effect of the antecedents to the outcome

variable as proposed by the model. Furthermore, Curcuruto and Griffin (2017) have

argued that safety knowledge is a distal antecedent of safety proactivity. However, the

high correlation found between safety knowledge and proactive safety behaviours in

the current study showed that it is possible that safety knowledge may act as a proximal

antecedent to proactive safety behaviours.

4.5.1 Practical Implications

The findings of the current meta-analysis provide practical value to

organisations. Specifically, the findings demonstrate that various organisational

antecedents have significant influences on one’s engagement with proactive safety

behaviours. Organisations are in the position to make changes within the workplace in

order to provide a space for their employees to feel safe in engaging in such

behaviours. The specific focus on organisational antecedents demonstrate that while a

selection process could distinguish workers who may have a disposition to act in a

proactive manner (e.g., promotion-focused), they can also be trained and encouraged

to take safety initiatives by cultivating a positive safety climate, supportive and caring

workplace, and designing a workplace that fosters autonomy.

The current review also highlighted that while safety climate is a critical factor

for safety proactivity, a variation exists depending on which level of safety climate is

being examined. The differentiation between the levels of safety climate allows for

more specific feedback to relevant parties (Zohar, 2005). In other words, by

distinguishing the levels of safety climate, organisations could have a specific focus

on which level to target for intervention. The current findings demonstrated that while

organisational-level safety climate is still a critical factor of proactive safety

behaviours, group-level safety climate is of particular importance as direct supervisors


and work groups may have more influence on proactive safety behaviours compared

to company-level safety procedures and policies. For instance, while Huang et al.

(2013) did not specifically studied proactive safety behaviours in work drivers, they

found that group-level safety climate has a stronger relationship with near-miss

incident reporting compared to organisational-level safety climate.

The current study also highlighted the different leadership styles that are critical

in encouraging proactive safety behaviours. A particularly unique contribution of this

study is demonstrating that having a passive leadership style could have a detrimental

impact on these behaviours. Therefore, leaders should take particular consideration of

not taking a passive approach when managing workers, as employees may be less

likely to be proactive towards workplace safety.

4.5.2 Limitations and Future Directions

Although the current meta-analysis may inform theory and practice, it is

important to mention its limitations. The measurement of proactive safety behaviours

was exclusively carried out with self-report measures and the majority were cross-

sectional studies. The use of self-report measures could introduce potential method

bias (Lindell & Whitney, 2001; Podsakoff, MacKenzie, Lee, & Podsakoff, 2003).

Furthermore, there is a limited number of peer-reviewed literature focused on work

driving safety, particularly on the topic of proactive safety behaviours. This limitation

prevented a comprehensive review of proactive safety behaviours within the context

of work driving.

Due to the small number of studies that both concurrently examined the

contextual and proximal antecedents within the same study, the current meta-analysis

could also not test a path analysis to model the extracted variables in order to determine

the possible mediating effects of the proximal antecedents. Future studies could apply

path modelling when more research becomes available on the topic of proactive safety

behaviours. The path analysis could further investigate the direction of the relationship

between the distal and proximal antecedents.

4.5.3 Chapter Summary and Key Learnings

The current chapter critically synthesised the contextual and proximal

antecedents of proactive safety behaviours and its related constructs. The review

specifically focused on safety climate, organisation-based social exchange and


interpersonal processes, and work design as contextual antecedents of proactive safety

behaviours while safety knowledge, regulatory focus, self-efficacy and perceived

control were considered as the proximal antecedents. The meta-analytic component of

the review extended upon Curcuruto and Griffin’s (2017) literature review on safety

proactivity and proposed model of safety proactivity. While Curcuruto and Griffin

(2017) provided a comprehensive account of the antecedents associated with safety

proactivity, the meta-analytic calculations allowed the pooling of available

quantitative data from individual studies for a more precise estimate of the effect size

of each relationship (Schmidt & Hunter, 2014). The current review was utilised to

inform the development of a research model of proactive safety behaviour that would

be applied within the work driving context.

74 Chapter 5: Development and Testing of Proactive Safety Behaviour Measure within the Work Driving Context

Chapter 5: Development and Testing of Proactive Safety Behaviour Measure within the Work Driving Context


The previous chapter discussed the terminologies, measures and concepts that

relate to proactive safety behaviours. Proactive safety behaviour is established as a

leading indicator of safety performance within the general OHS literature. However,

this concept is not yet examined within the work driving research despite several

researchers already indicating that this is an important concept to apply within the

work driving field (Otto et al., 2014; Newnam, et al., 2014; Wishart et al., 2019).

Furthermore, previous research on proactive safety behaviours (e.g., Curcuruto and

Griffin, 2016; Curcuruto et al., 2016) does not consider the unique challenges that

involved in work-related driving tasks (e.g., working alone, not being consistently

monitored by supervisors).

Within the OHS field, several measures exist that assess employees’ engagement

with proactive safety behaviours. However, as mentioned in the previous chapters,

most of these measures are unidimensional. Aside from Hoffman et al.’s (2003)

measure of safety citizenship, proactive safety behaviours are usually measured using

Neal and Griffin’s (2003) measure of safety participation which utilises 3 items. Other

versions of this survey include more items (e.g., Zacharantos et al., Subramaniam et

al., 2016) but these measures are still unidimensional in nature.

Similar concepts to proactive safety behaviours had been studied within the field

of road safety (e.g., traffic safety citizenship; Finley et al., 2015), as discussed in

Chapter 2. However, these studies only examined stewardship-type behaviours of

intervening when a driver is conducting an unsafe behaviour (e.g., using mobile phone

while driving or not wearing a seat belt; Buckley et al., 2009; Otto et al., 2016).

Thus, a research gap exists. There is a need to develop a measure that is specific

to the context of work driving as there might be certain elements within work driving

itself that may impact one’s proactivity to be safe. For instance, Newnam et al. (2011)

Chapter 5: Development and Testing of Proactive Safety Behaviour Measure within the Work Driving Context 75

developed the driving behaviour questionnaire scale within the occupational context

(O-DBQ) based on the commonly used general driving behaviour questionnaire

(DBQ). Newnam et al. (2011) argued that a specific scale was critical due to the

context of the workplace environment and its potential impact on employees’ driving

behaviours. They further argued that the scale could help in identifying unsafe driving

behaviours that are specific to the workplace context and could help organisations

monitor safe driving practices among their employees. Once the scale was developed,

Newnam and VonSchuckmann (2012) compared the predictive value of safety climate

and role overload between O-DBQ and DBQ and found that the organisational factors

accounted for a greater variance in the specifically tailored questionnaire compared to

the general DBQ.

Based on the similar premise, the current research program developed a measure

of proactive safety behaviours within the context of work driving (PSB-WD) to

consider the specific elements that are inherent in driving for work. For instance, it

was decided that items such as “trying to improve safety procedures” from Hofmann

et al.'s (2003) measure of safety citizenship should be altered to include “work driving

safety procedures” to indicate that the item specifically measures work driving safety.

Doing so improved the conceptual clarity and relevance of these items to suit the work

driving context rather than general occupational safety behaviours. It was also decided

that several items should be added to suit the scale within the work driving context.

For example, items specific to taking initiative with vehicle maintenance were added

to the measure to investigate proactive work driving practices.

The current chapter details the three studies that were conducted to develop a

multi-dimensional PSB-WD scale. This chapter has two sections: 1) the methods,

results, and discussion of Study 2a and 2b, which represents the scale development

and initial testing of the dimensionality of the scale, and 2) the methods, results, and

discussion of Study 2c which represents further testing of the scale. Table 5.1 shows

the overview of the current chapter. These studies aim to answer the second research

question: “What are the proactive safety behaviours that work drivers perform to

ensure and improve safety (and that of their co-workers) while driving for work?”


Table 5.1 Chapter section, phases, methodology and purposes of Study 2a, 2b and 2c

Chapter Section

Phase Method and Sample Purpose

Section 1: Scale development and initial testing of the scale dimensionality and reliability

Phase 1a: Item generation

(Study 2a)

Literature review

Research Team (N = 3)

Process akin to Group Nominal Technique

Generate an item pool for the survey

Phase 1b: Expert Panel

(Study 2a)

Expert Panel Survey

Sample of work drivers’ supervisors and managers and road safety experts (N = 5)

Assessment of content validity and clarity

Phase 2: Pilot testing of the items

(Study 2b)

Pilot Testing Survey; Principal Component Analysis

Sample of work drivers (N = 43)

Assessment of dimensionality and reliability, as well as item reduction

Section 2: Further scale testing

Phase 3: Testing of Survey Validity

(Study 2c)

Main Survey – Part A

Exploratory Factor Analysis

Confirmatory Factor Analysis (N = 300)

Further assessment of dimensionality and reliability, as well as construct validity, convergent validity and discriminant validity

5.2 PURPOSE OF THE STUDY

Careful consideration of specific aspects of work driving safety were taken into

account during the development process. A rigorous process was conducted to ensure

that the newly developed scale captured a variety of proactive safety behaviour

dimensions within the work driving context and to develop a psychometrically sound

scale. The following section details that process that was taken to develop, pilot and

validate a new measurement tool for proactive safety behaviours within the work

driving context. The procedure followed Hinkin’s (1998) and DeVellis (2016)

approaches to scale development.

5.3 METHOD AND RESULTS OF STUDY 2A – ITEM GENERATION AND EXPERT PANEL

According to Hinkin (1998), scale development typically follows two

approaches: deductive development and inductive development. In the deductive

approach, the theory informs the development of the scale (top-down). This approach


requires a thorough and extensive understanding of the phenomenon being studied in

order to inform the initial generation of the survey items (Hinkin, 1998). The inductive

scale development, on the other hand, does not follow a theoretical foundation and

requires researchers to use qualitative approaches in order to generate items for the

survey (Hinkin, 1998). For the current research, the deductive approach was utilised

due to the existing knowledge on proactive safety behaviours. The findings from the

literature review of the theoretical construct and assisted with the item generation.

Item Pool Generation

A literature review was conducted to inform the development of the theoretical

construct and to assist in developing the items for the survey. Part of this literature

review is evident in Chapter 2 of the thesis, which detailed the concepts, terminologies,

definitions and measures related to proactive safety behaviours. Various databases

were utilised to search for relevant literature and measures of safety citizenship, safety

participation, extra-role safety behaviours and safety proactivity. Other measures of

work contextual performance, organisational citizenship behaviours and work

proactivity were also reviewed due to the constructs origins in organisational

psychology research.

Items from existing measures were collected and reviewed (e.g., Geller et al.,

1996; Hofmann et al., 2003; Mearns & Reader, 2008; Neal et al., 2000; Parker &

Wang, 2015; Simard & Marchand, 1994; Willis, Brown, & Prussia, 2012). A total of

84 items were generated with 70 items derived from existing measures that were

modified to suit the work driving context and 14 items were originally developed.

Since the construct of proactive safety behaviour is multi-dimensional in nature,

the candidate, along with her supervisory team, utilised the Nominal Group

Technique15 to rank, revise, remove and categorise the pooled items into possible

dimensions of proactive safety behaviours. The research team were provided their own

15 Nominal Group Technique is a strategy used when a group is involved in a process of problem identification solution generation and decision making (Delbecq, Van de Ven, & Gustafson, 1975). When utilising this technique, group members provide their view of the solution with a short explanation. When duplicate solutions exist, these solutions are moved to the next phase of the solution generation and the members proceed to rank the remaining solutions (Delbecq et al., 1975). This technique encourages participation from all group members and results in a prioritised solution that represent the group’s preferences.


copies of the item pool and were given the task of assigning sub-dimensions for each

item, which included: whistle-blowing, stewardship, voice, improving safety,

problem-prevention, monitoring, feedback inquiry, taking charge, and safety civic

virtue. The dimensions were mostly based on Hofmann et al.’s (2003) research on

safety citizenship behaviours and dimensions of proactivity by Parker and colleagues

(Parker & Collins, 2010; Parker & Wang, 2015). Each researcher was also told to rank

the items from 0 (do not include) to 5 (include). Researchers were encouraged to

modify the item to improve its clarity and they were also encouraged to offer any

comments on each item.

If the research team could not provide a suitable category, they were instructed

to provide a new category for the item. If a disagreement between items occurred

(regarding the ranking or category), the research team had to provide reasons for their

ranking and assigned categories. Reasons for item modifications were also provided.

Disagreement in ranking and categories were discussed until consensus were reached

between the team members. For example, one of the team members suggested that

whistle-blowing may have harmful connotations as it may suggest aggressive

behaviours. All researchers agreed and this category was changed to voice, which was

considered as denoting a more neutral tone.

From this process, it was identified that proactive safety behaviours in the work

driving context could involve six key behavioural indicators, which are detailed below:

Problem prevention – acting to prevent the re-occurrence of challenges and

barriers to safety while driving for work.

Stewardship – helping co-workers to ensure their safety while driving for

work.

Voice – speaking up when employees have work driving safety concerns.

Helping / Volunteerism – volunteering to carry out activities that are not

formally part of their job as well as helping and cooperating with others to

ensure the safety of other employees while driving for work.

Feedback inquiry – explicit verbal requests for feedback regarding the

employees’ safety behaviours while driving for work.


Changing organisation’s policies and procedures – Taking one’s initiative

to change the organisational policies and procedures with an aim to

improve the safety of work drivers.

From the initial 84 items, 47 items remained in the item pool, with 5 items for

feedback inquiry, 5 items for volunteerism, 13 items for problem prevention, 10 items

for stewardship and 5 items for voice and 9 items for changing organisation's policies

and procedures. The 47 items were reviewed by the candidate and the supervisory team

to ensure the clarity of the items, and that they were not repetitive and thus distinct to

its assigned category. The scaling method used for this survey was a 5-point Likert

type scale ranging from 1 (never) to 5 (always) to determine the likelihood that work

drivers would perform these behaviours. This scaling method was utilised as it has

been employed in other surveys within the work driving settings (e.g., Newnam et al.,

2012).

Expert Panel

After developing the items for the survey, experts in fleet safety and occupational

safety research, as well as managers and supervisors of work drivers, were approached

to take part in an expert panel. The panel members were required to have either, a)

experience in managing fleets or supervising work drivers, or b) have experience in

research regarding work driving safety and occupational safety research. Potential

panel members were recruited from the professional network of the external supervisor

(Darren Wishart). An anonymous survey was developed using Qualtrics (a survey

management website) and links were sent via email. Online recruitment was preferred

due to the geographic dispersion of the potential panel members. Fifteen participants

were approached to participate and five participants completed the anonymous survey

online to examine the survey items (33.3% response rate).

The items were checked for the clarity of wording 1 (not clear) to 5 (very clear)

and content 1 (low content validity) to 5 (high content validity). Experts were also

requested to provide comments on each item and a general comment regarding the

dimensions. These comments were considered and changes were applied to the

existing items (e.g., minor changes in wording and sentence structure). See Appendix

B for the survey used.


Most items received good ratings of content validity and clarity. Due to the very

small sample size (n = 5), descriptives statistics were only conducted on the rating

scores. Items were either deleted or further modified if the content validity and clarity

ratings were very low. Some experts suggested that some behaviours within the item

pool were not possible for work drivers to carry out without approval from supervisors.

For example, for the item “I volunteer to conduct journey planning”, several experts

commented that work drivers need approval from their supervisors to conduct journey

planning and therefore, it was not applicable for work drivers. These items were

subsequently deleted.

Some experts also provided comments to improve the items. For example, a

expert commented that one of the items may seem aggressive due to the use of the

word “confront”. Therefore, some items were revised to reflect the comments provided

by the experts. From this process, five items were deleted and six items were revised:

Items Deleted:

I try to change the way the job is done so that I feel safe when driving for work

I volunteer to conduct work driving risk assessments

I volunteer to conduct journey planning

I try to prevent other drivers from being injured on the job

I resolve potential safety issues that relate to driving for work

I try to implement solutions to solve urgent work driving safety issue

Items Revised (added or edited phrases or words are italicised):

“I actively seek feedback about my work driving” to “I actively seek feedback

from my co-workers and passengers about my work driving”

“When I experience a traffic incident or receive a traffic offence when driving

for work, I initiate a conversation with people at work” to “When I experience

a traffic incident or receive a traffic offence when driving for work, I initiate

the conversation to learn from my mistakes”

“I get involved in work driving safety programs and activities even if it is

outside my responsibilities” to “I get involved in work driving safety programs

and activities to help other workers drive safely”


“I confront other co-workers about their unsafe behaviours while driving for

work” to “I would intervene if I see a co-worker doing something unsafe while

driving for work”

“I warn other co-workers about the dangers of driving unsafely” to “I

encourage my co-workers to think about the dangers of driving unsafely”

“I inform new drivers that violations of safety procedures will not be tolerated”

to “I inform other drivers that violations of safety procedures will have

negative consequences”

A total of 42 items remained. Due to the anonymity of the survey, the expert

panel did not have the opportunity to review the final items. However, these items

went further testing. The items were piloted using a sample of work drivers to

determine its factorability and, in the attempt to further reduce the number of the items.

The results of the pilot study are outlined in the following section.

5.4 METHOD AND RESULTS OF STUDY 2B – PILOT STUDY

Method

To establish the validity and reliability of the newly developed measure of

proactive safety behaviour within the work driving context, the questionnaire was

piloted with a State Government organisation that manages vehicle fleets. Data was

collected online using the Qualtrics survey platform. The link to the survey was sent

to the email of the organisation contact. Initially, 67 participants started the survey but

only 43 participants fully completed the study. Two participants indicated that they

were not eligible for the study, four participants stopped at the 3rd question and the

remainder of the incomplete data ranged from 2% to 87% completion. It is likely that

the length of the survey may have deterred the participants. The following results only

report data from the completed surveys based on N = 43 respondents. See Appendix C

and D for the participant information sheet and survey used for the pilot study.

Results

Participant Demographics and Work Driving Exposure

The majority of respondents were male (n = 23; 53.5%), aged 40 years and above

(n = 30; 69.2%) and had been employed by the company for more than 7 years (59.5%).

More than half (64.9%) of participants reported having held their driving licence for

more than 20 years. The majority of participants in the sample reported driving for


work between 1 to 10 hours per week (n = 38; 88.4%) and was driving up to 10,000

kilometres per year (n = 35; 81.4%).

Principal Component Analysis

A series of principal component analyses (PCA) was conducted to determine the

factor structure of the survey and identify poor performing (e.g., cross loading) items.

PCA is a technique used for reducing items and assessing possible dimensionality to

increase the measures interpretability but also minimising information loss (Jolliffe &

Cadima, 2016).

There were 42 items in the PSB-WD survey during the pilot study and 43

participants. A review on current developments in PCA technique demonstrated that

in studies that apply PCA with small sample sizes, especially in cases were n < p

(where n denotes the sample size and p denotes the number of variables or items), PCA

is still possible (Jolliffe & Cadima, 2016).

Furthermore, a Varimax rotation was applied because orthogonal rotations are

more desirable as they simplify the applied criteria and make the rotated factors easier

to interpret (Jolliffe & Cadima, 2016). On the other hand, non-orthogonal rotations

assume that factors are correlated but minimise factor loadings (Jolliffe & Cadima,

2016). Varimax is the most popular orthogonal rotation technique and commonly used

in social sciences (Jolliffe & Cadima, 2016). Even though previous literature suggests

that dimensions of proactive safety behaviours were correlated (e.g., Curcuruto,

Conchie, Mariani, & Violante, 2015), an orthogonal rotation was chosen over non-

orthogonal for ease of interpretation of the factors. The current study was only an initial

pilot-testing of the items and, therefore, the ability to interpret the factors was essential.

Next, the Bartlett’s test of Sphericity and Kaiser–Meyer–Olkin (KMO) Measure

of Sampling Adequacy were assessed to determine whether sufficiently large

relationships existed within the sample data to perform PCA and the subsequent

Exploratory Factor Analysis (EFA) in Study 2c (Howard, 2016). Bartlett’s test of

sphericity evaluates whether the sample’s correlation matrix is an identity matrix –

identity matrix indicates that the variables are unrelated and, therefore, not suitable for

structure detection (Howard, 2016). Given that PCA determines the variables’

relationships, a lack of correlations within the observed data prevents the PCA from

being performed (Howard, 2016). A significant Bartlett’s test of sphericity indicates

that the observed data is not an identity matrix and that PCA is appropriate.


However, several researchers have noted that the Bartlett’s test of sphericity is

usually significant as virtually most datasets are significantly different from an identity

matrix (Howard, 2016). Therefore, KMO Measure of Sampling Adequacy is usually

interpreted alongside Bartlett’s test of sphericity. KMO Measure of Sampling

Adequacy evaluates whether the proportion of variance within the observed data may

indicate latent factors. High values (close to 1) indicate higher inter-correlations within

the variables and values less than 0.50 are usually indicative that the EFA is not

suitable for the observed dataset.

For the initial PCA, a significant Bartlett’s test (2 (561) = 1535.30, p < .001)

was found and the KMO test of sampling adequacy was .38. While the KMO test of

sample adequacy was poor, the significant Bartlett’s test of sphericity suggested that

there may be significant inter-relationships between the variables and thus that the

factor model was appropriate (Hair, William, Babin, & Anderson, 2010; Tabachnick

& Fidell, 2007). The items for ‘volunteerism’ and the items for ‘changing of

organisational policies and procedures’ seemed to be converging into a single factor

structure. The research team decided to remove the items that had been created for

‘changing of organisational policies and procedure’ given that this factor may be more

likely to act as a higher factor structure that encompasses other sub-factors.

Initially, 8 factors were extracted from the initial PCA. After the removal of the

‘changing of organisational policies and procedure’ factor, items were deleted one by

one, removing highly-correlated items, high cross-loadings and low communalities.

Items were retained if they had strong factor loadings, small cross-loadings across all

factor structures and if the item provided a meaningful and useful contribution within

the factor. The final factor analysis still demonstrated 8 components with an

Eigenvalue of >1 (See Table 5.2). The KMO was .67, which is acceptable, and the

Bartlett’s test of sphericity was significant, 2 (325) = 1112.03, p < .001.


Table 5.2 Final PCA for the Pilot Study Survey PSB-WD

FACTOR LOADINGS

ITEM CODE

ITEM STATEMENT 1 2 3 4 5 6 7 8

Factor 1 – Helping / Volunteerism (4 items)

H/V5 I volunteer to help other drivers with their work safety driving responsibilities.

0.93

H/V4 I volunteer to help other workers learn more about safe work driving practices.

0.91

H/V3 I get involved in work driving safety programs and activities to help other workers drive safely.

0.86

H/V 1 I volunteer to educate work driving safety procedures to new drivers.

0.80

Factor 2 - Protecting Other Drivers and Fixing Safety Issues Outside of Responsibility (4 items)

S8 I go out of my way to look out for the safety of other drivers

0.85

S9 I take action to protect other drivers from risky situations.

0.84

PP6 If I see something unsafe, I go out of my way to take care of it.

0.72

PP7 I fix safety issues that relates to work driving even if it is not my responsibility.

0.60 0.45

Factor 3 - Intervening Other Drivers and Problem Prevention (5 items)

S3 I would intervene to stop safety violations of other drivers.

0.78

S2 I would intervene if I see a co-worker doing something unsafe while driving for work.

0.77 0.40

PP1 When driving for work, I plan extra journey time and breaks for bad weather, traffic congestion, etc.

0.75

PP3 I resolve problems in ways that reduce the risks associated with driving for work.

0.58 0.49

S1

If I see my co-workers doing something risky while driving for work (e.g., using their mobile phone or speeding), I talk to them about the hazards of their risky behaviours.

0.43 0.54

Factor 4 - Voice (3 items)

V4 I speak up on work driving safety matters even if others might disagree.

0.88

V5 I communicate my views about work driving safety issue, even if others would disagree.

0.84

V3 I speak up about safety concerns during team meetings or toolbox talks.

0.81

Factor 5 - Vehicle Maintenance (2 items)

PP8 If I notice a defect in the vehicle I am driving, I take an appropriate action by notifying my supervisors or completing paperwork.

0.90

PP9 When I see a vehicle that needs maintenance, I inform my supervisors or the appropriate person about it.

0.89

Factor 6 – Sharing Knowledge to Other Drivers (3 items)

S6 I encourage other drivers to follow safe work driving procedures.

0.85

S7 I encourage new drivers to follow safe working procedures.

0.40 0.78

S5 I inform other drivers that violations of safety procedures will have negative consequences.

0.76


FACTOR LOADINGS

ITEM CODE

ITEM STATEMENT 1 2 3 4 5 6 7 8

Factor 7 - Feedback Inquiry (3 items)

FI1 I actively seek feedback from my supervisor about my work driving.

0.87

FI2 I actively seek feedback from my co-workers and passengers about my work driving.

0.86

FI3 I actively seek feedback about my work driving with people at work.

0.76

Factor 8 - Fixing Safety Issues (2 items)

PP4 I implement solutions to solve safety issues that relate to driving for work.

0.87

PP5 When I see a potential work driving safety hazard, I do my best to fix it.

0.50 0.56

Total Variance Extracted 35.8 12.9 9.7 7.6 6.3 5.1 4.8 3.9

Internal Reliabilities .95 .89 .81 .95 .98 .94 .80 .74

Notes. Extraction Method: Principal Component Analysis; Rotation Method: Varimax with Kaiser Normalisation; Rotation converged in 13 iterations; N = 43. H/V = Helping/Volunteer Items; PP = Problem Prevention Items; FI = Feedback Inquiry; Stewardship; V = Voice.

As shown in Table 5.2, all extracted factors had reliability scores above

Cronbach’s α of 0.70, which is an acceptable measure for a scale’s internal consistency

(Kline, 2000). Cumulatively, the overall factor structure accounted for 86.1% of the

total variance. However, the results revealed that some items still loaded highly across

more than one factor. In order to maintain a minimum of three items per factor, these

items were still retained but were later modified to more clearly represent their primary

factor (Hinkin, 1998).

It was also noted that some of the factor structure seemed to represent two

different sub-factors. For instance, the items in Factor 2 seemed to represent

“Protecting Other Drivers” and “Fixing Safety Issues Outside of Responsibility” while

Factor 3 seemed to represent “Intervening Other Drivers” and “Problem Prevention”.

These factors were considered as representing two different categories of proactive

safety behaviours.

In order to maintain at least three items per factor (to maximise reliability),

several additional items were written for factors that had less than three items. Another

item that represented protecting other drivers was added as well as an item reflecting

fixing safety issues outside of one’s responsibility. Similarly, another item was

developed for Factor 5 (Vehicle Maintenance). In addition, the items for the remaining

volunteering dimension were further revised to denote a more proactive stance on the

behaviour. More specifically, instead of “I volunteer…”, the items were revised to: “I


go out of my way…” The following table presents the revised and additional items to

the survey.

Table 5.3 Revised and added items

Revised Items

If I see my co-workers doing something risky while driving for work (e.g., using their mobile phone or speeding), I talk to them about the hazards of their risky behaviours. Revised to: If I see my co-workers doing something risky while driving for work (e.g., using their mobile phone or speeding), I would intervene. I resolve problems in ways that reduce the risks associated with driving for work.

Revised to: When I experience safety issues while driving for work, I resolve these problems. When I see a potential work driving safety hazard, I do my best to fix it. Revised to: When I see a potential safety hazard while driving for work, I resolve it by finding a solution. Helping/Volunteer items were revised to: I go out of my way to educate new employees about work driving safety procedures. I go above my duties to help other drivers with their work driving safety responsibilities. I go out of my way to help other drivers learn more about safe work driving practices. I go above my duties to help my co-workers drive safely.

Additional Items

I go above my duties to resolve driving safety issues at work for Fixing Safety Issues I make sure that vehicle defects are attended to quickly for Vehicle Maintenance I look after the safety of other drivers in my organisation for Stewardship Protection

5.5 SUMMARY OF STUDY 2A AND 2B

Study 2a and 2b presented the scale development process for the PSB-WD

measure. Study 2a involved the item pool generation and expert panel review of the

items, while Study 2b involved the pilot testing of the measure. From the item

generation phase, the process generated 84 items that were reduced to 47 items using

the Nominal Group Technique decision process applied by the research team. The

generated item pool was categorised into six dimensions of proactive safety behaviours

within the work driving context, namely: feedback inquiry, helping/volunteerism,

problem prevention, stewardship, voice and changing organisational policies and

procedures. Then, an expert panel was consulted to further review and give feedback

on the items. From this process, the measure was further reduced to 42 items (with six

revised items).

Although the pilot study had a small sample size and the results from the PCA

should be interpreted cautiously, the findings were able to provide important insights

into the dimensionality and reliability of the PSB-WD scale as well as potential


improvements. It was hypothesised that the proactive safety behaviours scale within

the work driving context (PSB-WD) may have six factors (see Figure 5.1). The results

from the initial PCA indicated that changing of organisational policies and procedures

may be indicating a higher-order factor and, therefore, this dimension was removed

from subsequent PCA. The final PCA suggested that, generally, most items remained

within the hypothesised factors. For instance, the items developed for

volunteerism/helping, voice and feedback inquiry were originally written for these

factors.

However, the findings from the pilot study also revealed that some of the

extracted factors could be further divided into sub-dimensions. In particular, it is

possible that problem prevention could be divided into fixing responsibilities and

vehicle maintenance, while stewardship could be divided into protection, intervention

and knowledge-sharing. Several new items were developed to ensure that all

hypothesised sub-factors had three items, as recommended by Bollen (1989, p. 244),

so as to ensure stability of each factor. Therefore, the revised PSB-WD resulting from

Study 2 had 29 items. The following section will discuss the evaluation of the validity

of the factors and the overall measure.

PSB-WD

Study 2b

Feedback Inquiry

Helping / Volunteerism

Problem Prevention

Fixing safety issues

Fixing safety issues outside of

responsibility

Vehicle maintenance

Stewardship

Protection

Intervene

Sharing KnowledgeVoice

PSB-WD

Study 2a

Feedback Inquiry


Problem Prevention

Stewardship

Voice

Changing of organisations policies

and procedures

Figure 5.1 Hypothesised dimensions of proactive safety behaviours within the work driving context (PSB-WD; left) and the dimensions extracted from PCA using the pilot study data (right).


5.6 METHOD AND RESULTS OF STUDY 2C

5.6.1 Method

Study 2c was conducted to further assess the dimensionality of the PSB-WD

scale, as well as further assess its reliability and validity. An online survey was

primarily used to access a wider range of employees who drive for work. Potential

participants for Study 2c were informed that the current study was interested in

understanding the proactive behaviours that work drivers and their supervisors perform

to ensure employees’ safety while driving for work. The anonymous questionnaire,

developed using Qualtrics, took 25 to 30 minutes to complete. The recruitment

methods included: convenience sampling via snowballing, a social media campaign

via Facebook, accessing members of various fleet management associations (where

members manage small to large vehicle fleets from 10 to 10,000 fleets) and email

advertisements. The online recruitment yielded N = 219 usable surveys (with less than

10% missing data). See Appendix E and F for the participant information sheet and

survey used for Study 2c.

In addition, hardcopies of the survey were distributed during a series of

workshops delivered among work drivers who were employed within the State

Government and the workshops were run by the external supervisor (Darren Wishart).

There were 119 participants who attempted the hardcopy version of the survey.

However, 38 of the hardcopy surveys showed substantial missing data. Therefore, only

81 surveys were used. Input of hardcopy data were validated by a random selection

(30%) of the hardcopy surveys and were crosschecked with the entered data. Data

entries that were found to be inaccurate were subsequently corrected.

To maximise the participation rate, participants who completed the survey had

the chance to go into a random prize draw to win one of ten $100 gift vouchers. After

completing the survey, participants who wanted to be included in the prize draw were

provided with a separate link to enter their personal details so as to keep such details

separate from their survey responses and thus ensured their anonymity.

5.6.2 Results of Study 2c

Overview of Data Analysis

To assess the validity and reliability of the newly developed scale, EFA and

Confirmatory Factor Analysis (CFA) were performed. EFA was conducted in order to


identify the underlying dimensions among the set of variables and to assess whether

the resulting factors corresponded to the hypothesised dimensions (Hair et al., 2010).

CFA, on the other hand, quantifies the scale validity and the goodness-of-fit of the

resulting model from the EFA (Hair et al., 2010). The goodness-of-fit is the extent to

which the sample data matches the expected values from the hypothesised model. All

data analyses were conducted in IBM SPSS Statistics version 23 and IBM AMOS

version 25.

Assessment of Missing Data

Data analyses were carried out on usable surveys where cases had missing data

on less than 10% of their responses. There were 300 usable cases in the dataset. There

were 219 surveys completed online and 81 on paper.16 As recommended by Hair et al.

(2010), missing data under 10% for individual cases can generally be ignored (p. 55).

From the remaining surveys, analysis of the raw data revealed that there were minimal

missing information across variables (0.3% to 2.0%). the Missing Value Analysis

(MVA) was carried out and the non-significant Little’s MCAR test, χ2 (2353) =

2164.68, p = .997 suggesting that the data were missing completely at random.

Participant Demographics and Work Driving Exposure

Table 5.4 presents the information on the demographic and work characteristics

of the sample. The majority of respondents were work drivers who did not have a

supervision role, had an average age of 36 years (SD = 14.21; Range = 17 to 74), and

who had held their driving licence for an average of 18 years (SD = 14.07; Range =

less than one year to 56 years). The participants had been working within their current

organisation for an average of 7 years (SD = 8.24; Range = 0 to 40 years). More than

half (55.7%) of the participants reported driving for work between 1 to 10 hours per

week and almost half (42.7%) reported to drive between 1 to 10,000 kilometres for

work annually. Most participants reported driving a car, sedan or a wagon type of

vehicle for work (59.7%) and that they drove on asphalt or bitumen roads (84.3%).

The majority of the sample (87%) also reported not having previously been involved

in a crash and 83.3% reported that they had not received a traffic offence in the past

16 EFA for both the online and hardcopy version of the survey were assessed and no significant differences were found regarding factor structures and item loadings for each structure.


year. However, around 62.3% reported having experienced at least one near-miss in

the past year.

Table 5.4 Demographic and work information of participants

Frequency Percent Data Collection Method Online Survey 219 73.0 Hardcopy Survey 81 27.0 Supervision Role Yes 32 10.7 No 268 89.4 Work driving hours per week 1 – 10 hours 167 55.7 11 – 20 hours 75 25.0 21 – 30 hours 28 9.3 31 – 40 hours 16 5.3 41 – 50 hours 6 2.0 51 – 60 hours 5 1.7 61 hours or more 3 1.0 Annual kilometres driven for work 1 – 10,000 kms 128 42.7 10,001 - 20,000 kms 68 22.7 20,001 - 30,000 kms 38 12.7 30,001 - 40,000 kms 22 7.3 40,001 - 50,000 kms 17 5.7 50,001 - 60,000 kms 10 3.3 60,001 kms or more 17 5.7 Gender Male 148 49.3 Female 152 50.7 Type of vehicle for work Car / Sedan / Wagon 179 59.7 SUV / 4WD 83 27.7 Heavy Vehicle 20 6.7 Other 18 6.0 Type of road Asphalt / Bitumen roads 253 84.3 Dirt roads 3 1.0

Combination of Asphalt / Bitumen roads and Dirt roads

41 13.7

Other 2 0.7 Missing 1 0.3 Self-reported crash in the past 12 months

None 261 87.0 One crash 29 9.7 Two crashes 6 2.0 Three or more crashes 4 1.3 Self-reported near misses in the past 12 months

None 113 37.7 One near-miss 73 24.3 Two near-misses 48 16.0 Three or more near-misses 66 22.0 Self-reported traffic offences in the past 12 months

None 250 83.3 One traffic offence 29 9.7 Two traffic offences 10 3.3 Three or more traffic

offences 11 3.7

Notes. N = 300.


5.6.3 Exploratory Factor Analysis

Assumption Checking

Several assumptions were checked to ensure that the data was suitable for the

factor analysis. First, the current sample size of N = 300 was an appropriate number

for factor analysis, as recommended by several authors such as Hair et al. (2010) and

Tabachnick and Fidell (2007). Next, the Bartlett’s test of Sphericity and Kaiser–

Meyer–Olkin (KMO) Measure of Sampling Adequacy were assessed to determine

whether there were sufficiently large inter-relationships within the sample data to

perform EFA (Howard, 2016).

Factorability of the correlation matrix was also examined. A correlation matrix

should be used in the EFA process to display the relationships between individual

variables. Tabachnick, Fidell, and Osterlind (2007) recommended inspecting the

correlation matrix for correlation coefficients over 0.30. Hair et al. (2010), on the other

hand, suggested that if the loadings of ±0.30 is considered minimal, while ±0.40 is

important and ±.50 is significant.

Initial Exploratory Faction Analysis

For the EFA, a series of Principal Axis Factoring (PAF) analyses was conducted

to determine the factor structure of the scale. While the previous pilot testing of the

survey used an orthogonal rotation to simplify the rotations for the PCA, an oblique

(correlated) rotation was carried out for PAF as it allows the factors to be correlated.

For the current study, Promax Rotation, which is an oblique rotation, was chosen as it

has the advantage of being conceptually simple and offers fast calculations (Lewis-

Beck, Bryman & Liao, 2003). Furthermore, the pilot study that was previously

conducted utilised PCA with a Varimax Rotation. Using Promax allowed for

continuity with previous analysis, given that Promax utilises a target matrix usually

obtained from a Varimax Rotation for the factor rotation (Abdi, 2003).

Examining the items’ correlation matrix, almost all items had significant inter-

item correlations. Only two inter-item correlations were non-significant: volunteer

item 4 with vehicle maintenance item 1, r = 0.06, p = .135 and volunteer item 4 with

vehicle maintenance item 2, r = 0.05, p = .220. All other inter-item correlations were

significant, ranging from r = 0.10 to r = .85. An initial PAF with Promax Rotation was

conducted and factorability of the overall set of variables was assessed. The KMO

measure of sampling adequacy was .93 and a significant Bartlett’s Test of Sphericity,


2 (406) = 6673.98, p < .001. The large inter-item correlations, the high KMO and

significant Barlett’s Test of Sphericity suggested that the data were suitable for EFA.

Having assessed the assumptions of the EFA and established the suitability of

the observed sample data for the analysis, the 29 item PSB-WD scale was subjected to

a PAF with Promax Rotation. The initial PAF revealed six factors with eigenvalues

exceeding 1, explaining 66.13% of the total variance (See Table 5.5). The

communalities and pattern matrix were examined to assess the factor loadings and to

identify any items that have low communalities and/or which had cross-loadings with

multiple factors. From these examinations, the initial PAF results demonstrated that

there were items with a) low communality (n = 1) and b) high cross-factor loadings (n

= 4). Items with high cross-factor loadings and low communalities were removed and

the factor model was re-specified to improve the simplicity of the factor structure.

Furthermore, inter-item correlations were also assessed for highly correlated items as

it may indicate item redundancy. These items were deleted one-by-one to assess how

the deletion would impact on the overall factor structure.


Table 5.5 Initial PAF with Promax Rotation

FACTOR LOADINGS ITEM CODE ITEM STATEMENT 1 2 3 4 5 6 V 3 I go out of my way to help other drivers learn more about safe work

driving practices. 0.97

V 2 I go above my duties to help other drivers with their work driving safety responsibilities.

0.88

V 4 I go above my duties to help my co-workers drive safely. 0.82

SK 1 I encourage new drivers to follow safe working procedures. 0.82

V 1 I go out of my way to educate new employees about work driving safety procedures.

0.78

SK 3 I encourage other drivers to follow safe working procedures. 0.76

SK 2 I inform other drivers that violations of safety procedures will have negative consequences.

0.57

SP 3 I look after the safety of other drivers in my organisation.e 0.47

SP 1 I take action to protect other drivers from risky situations.b 0.30 0.29 0.24 0.27 -0.20 PP-G 2

I implement solutions to solve safety issues that relate to driving for work.

0.90

FIS 1 I fix safety issues that relate to work driving even if it is not my responsibility.

0.86

PP-G 3

When I see a potential safety hazard while driving for work, I resolve it by finding a solution.

0.79

PP-G 1

When I experience safety issues while driving for work, I resolve these problems.

0.67

FIS 3 I go above my duties to resolve driving safety issues at work.c 0.30 0.67

FIS 2 If I see something unsafe, I go out of my way to take care of it.f 0.54

PP-G 4

When driving for work, I plan extra journey time and breaks for bad weather, traffic congestion, etc.a

-0.26 0.53

SP 2 I go out of my way to look out for the safety of other drivers.d 0.22 0.27 0.25

SI 2 If I see my co-workers doing something risky while driving for work (e.g., using their mobile phone or speeding), I would intervene.

0.98

SI 1 I would intervene if I saw a co-worker doing something unsafe while driving for work.

0.87

SI 3 I would intervene to stop safety violations of other drivers. 0.69

FI 2 I actively seek feedback from my co-workers and passengers about my work driving.

0.91

FI 3 I ask people at work for feedback about my driving. 0.85

FI 1 I actively seek feedback from my supervisor about my work driving. 0.73

VM 2 When I see a vehicle that needs maintenance, I inform my supervisors or the appropriate person about it.

1.00

VM 1 If I notice a defect in the vehicle I am driving, I take an appropriate action by notifying my supervisors or completing the paperwork.

0.91

VM 3 I make sure that vehicle defects are attended to quickly. 0.57

V 2 I speak up about work driving safety issues that affect work drivers even if it makes me unpopular.

0.83

V 3 I communicate my views about work driving safety issue, even if others would disagree.

0.69

V 1 I speak up about safety concerns during team meetings or toolbox talks. 0.69 Total Variance Explained 42.45 8.79 4.48 3.95 3.59 2.87 Notes. Extraction Method: Principal Axis Factoring with Promax Rotation; Rotation converged into 7 iterations. a Problem prevention item 4 was deleted due to low communality (.27) b Steward protection item 1 was deleted due to high cross-factor loadings. c Fixing responsibility item 3 was deleted due to high cross-factor loadings. d Steward protection item 2 was deleted due to high cross-factor loadings. e Steward protection item 3 was deleted due to high cross-factor loadings. The initial factor analysis did not show cross-factor loadings, but cross-loadings were visible once the items were deleted one-by-one. f Fixing responsibility item 2 was deleted due to high cross-factor loadings. The initial factor analysis did not show cross-factor loadings, but cross-loadings were visible once the items were deleted one-by-one. SK = Steward Knowledge; V = Volunteer; V = Voice; VM = Vehicle Maintenance; FI = Feedback Inquiry; SI = Steward Intervene; PP = Problem Prevention - General; Fixing Safety Issues = FSI


The initial PAF showed that Problem Prevention General item 4 had a low

communality score of .27 after rotation. This item was deleted first and the PAF was

re-run. Steward Protection item 1 showed high cross-loadings with multiple factors

and was removed from the analysis. Fixing Safety Issues item 3 also demonstrated

high cross-factor loadings. The initial PAF did not indicate that Steward Protection

item 3 and Fixing Safety Issues item 2 had high-cross factor loadings, but the cross-

loadings were visible on the 4th and 5th re-run of the analysis. All items for Steward

Protection were removed. Overall, 6 items were removed from the EFA procedure.

Final Exploratory Factor Analysis

The final PAF revealed the same 6 factors, with a total of 70.2% variance

explained. Inspection of the item statements for each factor inferred that the first factor

represented “education”, factor two represented “feedback inquiry”, factor three

represented “intervene”, factor four represented vehicle maintenance, factor 5

represented “fixing safety issues” and lastly, factor 6 represented “voice”. Examination

of the Cronbach’s alpha for the internal consistency of the items showed that all factors

were above 0.80 which indicates an acceptable reliability (Kline, 2000) and well above

the minimum acceptable guidelines of 0.70 for newly developed scales (DeVellis,

2016).

Inspection of the factor correlation matrix showed that the strength of the

relationships between the factors was generally medium, ranging between r = 0.25 to

r = 0.68, except for the low correlation between feedback inquiry and vehicle

maintenance (r = 0.13). Thus, using Promax, which is an oblique factor rotation, was

appropriate for the analysis.


Table 5.6 Final PAF with Promax Rotation

FACTOR LOADINGS ITEM CODE ITEM STATEMENT 1 2 3 4 5 6 Factor 1 - Educate

Volunteer Q3 I go out of my way to help other drivers learn more about safe work driving practices.

0.93

Volunteer Q2 I go above my duties to help other drivers with their work driving safety responsibilities.

0.84

Steward Knowledge Q1

I encourage new drivers to follow safe working procedures. 0.80

Volunteer Q4 I go above my duties to help my co-workers drive safely. 0.80


I encourage other drivers to follow safe working procedures.

0.75

Volunteer Q1 I go out of my way to educate new employees about work driving safety procedures.

0.75


I inform other drivers that violations of safety procedures will have negative consequences.

0.61

Factor 2 – Fixing Safety Issues Problem Prevention 2


0.84

Problem Prevention 3


0.80

Problem Prevention 1


0.73

Fixing Responsibility Q1

I fix safety issues that relate to work driving even if it is not my responsibility.

0.67

Factor 3 – Feedback Inquiry

Feedback Q2 I actively seek feedback from my co-workers and passengers about my work driving.

0.94

Feedback Q3 I ask people at work for feedback about my driving. 0.87

Feedback Q1 I actively seek feedback from my supervisor about my work driving.

0.76

Factor 4 – Intervene

Steward Intervene Q2

If I see my co-workers doing something risky while driving for work (e.g., using their mobile phone or speeding), I would intervene.

0.95


I would intervene if I saw a co-worker doing something unsafe while driving for work.

0.81


I would intervene to stop safety violations of other drivers. 0.65

Factor 5 –Voice

Voice 2 I speak up about work driving safety issues that affect work drivers even if it makes me unpopular.

0.94

Voice 3 I communicate my views about work driving safety issue, even if others would disagree.

0.81

Voice 1 I speak up about safety concerns during team meetings or toolbox talks.

0.79

Factor 6 – Vehicle Maintenance

Vehicle Maintenance Q2

When I see a vehicle that needs maintenance, I inform my supervisors or the appropriate person about it.

0.99


If I notice a defect in the vehicle I am driving, I take an appropriate action by notifying my supervisors or completing the paperwork.

0.84


I make sure that vehicle defects are attended to quickly. 0.51

Total Variance Explained % 42.9 9.6 5.4 4.9 4.0 3.3 Cronbach Alpha 0.93 0.90 0.87 0.92 0.86 0.84 Notes. Extraction Method: Principal Axis Factoring; Rotation Method: Promax with Kaiser Normalisation. Rotation converged in 7 iterations.


5.6.4 Confirmatory Factor Analysis

Overview of Confirmatory Factor Analysis

For the CFA, the model’s goodness of fit was assessed using three indices:

absolute fit, incremental fit and parsimony fit. Absolute fit indices determine the extent

that a proposed model fits the observed sample data and identified which of

hypothesised model provided the most superior fit (Byrne, 2016). The calculation for

these absolute fit indices does not rely on a comparison with a baseline model, instead

it measures how well the observed model compares to a non-existent model (Byrne,

2016). Examples of absolute fit indices include: Chi-square test, Goodness of Fit Index

(GFI), Adjusted Goodness of Fit Index (AGFI), Root Mean Square Error of

Approximation (RMSEA) and the Standardised Root Mean Square Residual (SRMR).

Incremental fit indices, on the other hand, compare the hypothesised model to a

baseline model (Brown, 2014). Also known as comparative or relative fit indices, the

baseline model or the “null hypothesis” rests on the premise that covariances or

correlations among the indicators are fixed to zero (Brown, 2014). Comparative Fit

Index (CFI) and Tucker-Lewis Index (TLI) belong to this fit indices category.

Lastly, parsimony fit indices favour the models that are simple and have a better

model parsimony (i.e., number of freely estimated parameters expressed by model

degrees of freedom; Brown, 2014). Models that are less parsimonious and quite

complex are penalised – the more complex the model is, the lower the fit index.

Parsimony Goodness-of-Fit Index (PGFI), the Parsimonious Normed Fit Index (PNFI)

and CMIN/DF or Minimum Discrepancy divided by Degrees of Freedom are some

examples of parsimony fit indices (Byrne, 2016).

Table 5.7 shows the recommended threshold for each index. The ideal model

should not only be theoretically sensible, but must also fit the sample data. There is an

on-going discussion as to which indices are the most appropriate to report and what

the recommended thresholds should be (Brown, 2014; Byrne, 2016). For instance, the

conventional overall test of fit in covariance structure analysis is the chi-square (χ2)

test, which measures the extent of the discrepancy between the fitted covariance matrix

with the sample data (Hu & Bentler, 1995). A good model fit provides a non-

significant result at a 0.05 threshold (Brown, 2014). However, the χ2 test is quite

restrictive as it assumes multivariate normality and is quite sensitive to sample size. If

the sample size is large, the χ2 test usually rejects the model but a small sample size


could lead to a lack of power. Therefore, it is difficult to discriminate between good-

fitting from bad-fitting models using the χ2 test. The χ2 statistic should always be

interpreted in conjunction with other model fit indices (Hair et al., 2010). To assess

the model of the PSB-WD scale, the goodness of fit were assessed using Chi-square,

RMSEA (absolute fit indices), CFI (incremental fit indices) and CMIN/DF

(parsimonious fit indices). The SRMR was not calculated due to some missing data.

Table 5.7 Assessment of Model Fit

Criterion Excellent Threshold

Acceptable Threshold

Absolute Fit Indices

χ2 p 0.05 p 0.05 RMSEA < . 05 < . 08 SRMR < . 05 < . 08 GFI > .95 > .90 Incremental Fit Indices

NFI > .95 > .90 TLI > .95 > .90 RNI > .95 > .90 CFI > .95 > .90 Parsimony Fit Indices

AGFI > .95 > .90 CMIN/DF > .95 > .90 PNFI > .95 > .90

Notes. χ2 = Discrepancy Chi-square, CMIN/DF = Minimum Discrepancy/Degree of Freedom, RMSEA = Root Mean Square Error of Approximation, SRMR = Standardised Root Mean Residual, GFI = Goodness-of-Fit-Index, TLI = Tucker-Lewis Index.

Results of the Confirmatory Factor Analysis

Overall, the EFA analysis revealed that the PSB-WD scale had six factors:

Education, Feedback Inquiry, Intervene, Vehicle Maintenance, Fixing Safety Issues

and Voice. In order to confirm the measurement model and to assess its construct

validity, CFA was applied (Hair et al., 2010).

The initial model’s goodness of fit was 2 (215) 559.65, p < .001, CMIN/DF =

2.60, CFI = .93, RMSEA = 0.07. Although the 2 was significant, the other indices

suggested that the model was an acceptable fit to the data. Inspection of the results

revealed that the standardised loading estimates of all items were .60 or greater and

significant at p <.001, confirming the factor structure specified in the EFA.

The modification indices with complete data were checked to see if the model

could be re-specified to improve the model fit. The modification indices demonstrated


that the model could be improved by co-varying the error terms between steward

knowledge item 3 and item 1 which would reduce the largest modification in the model

(MI = 109.87). In addition, error terms for volunteer item 2 covaried with steward

knowledge item 1 (MI = 20.49), item 2 (MI = 17.03) and item 3 (MI = 19.26). While

various researchers have suggested the need to exercise caution if using the

modification indices for model re-specification, if the change is based on conceptual

terms, then the modification indices could be a useful tool (Byrne, 2016). For the

proposed change, the two items were conceptually belonging within the same factor

and, on reading the items description suggested that the two items were very similar

to each other, which could indicate a redundancy between the two items. Therefore,

instead of co-varying the two error terms, steward knowledge item 1 was deleted and

the model produced a better goodness-of-fit indices. Steward knowledge item 1 was

deleted as it reduced the reliability estimates and this item had referred to new drivers

while item 3 had referred to other drivers. Thus, the term “other drivers” may seem

more general and have broader appeal. Volunteer item 2 was also removed from the

analysis due to potential item redundancy with other volunteer items.

Figure 5.2 shows the re-specified measurement model of PSB-WD scale after

the removal of steward knowledge item 1 and volunteer item 2. The goodness-of-fit of

the specified model was 2(174) = 348.75, p < .001, CMIN/DF = 2.00, CFI = .96,

RMSEA = 0.06. The 2 test was still significant, but the other indices demonstrate a

better fit compared to the previous model.


Figure 5.2 Re-specified measurement model of PSB-WD scale

Validity Assessment

Validity of the measure was further explored by examining its construct,

convergent and discriminant validity (Hair et al., 2010). Convergent validity is a

function of the association between two different measurement scales, which are

supposed to measure the same concept and is achieved when multiple indicators

operate in a consistent manner (Hair et al., 2010). The composite reliability (CR) and

average variance extracted (AVE) were used to measure the convergent validity of the

constructs. The constructs are said to have convergent validity when the composite

reliability exceeds the criterion of 0.07 and the AVE 0.05. AMOS does not produce


the AVE and composite reliability for each factor, therefore, they were calculated

manually using an excel tool developed by Gaskin (2011) .

Discriminant validity is the extent to which the scales reflect their suggested

construct different from the relationship with all the other scales in the research model

(Hair et al., 2010). To test discriminant validity, the square root of AVE should be

higher than the correlations between the factors in order to satisfy the discriminant

validity requirement. As shown in Table 5.8, all square roots of the AVEs (in diagonal

cells) are higher than the correlations between the factors and demonstrating support

for discriminant validity.

Table 5.8 Discriminant and convergent validity

CR AVE Voice Educate Feedback Intervene

Vehicle Mainte- nance

Fixing Safety Issues

Voice 0.92 0.80 0.89

Educate 0.90 0.65 0.70 0.81

Feedback 0.90 0.76 0.39 0.61 0.87

Intervene 0.86 0.68 0.45 0.49 0.25 0.82

Vehicle Maintenance 0.85 0.66 0.45 0.37 0.16 0.42 0.81

Fixing Safety Issues 0.87 0.62 0.63 0.66 0.35 0.50 0.55 0.79 Notes. CR = Composite Reliability; AVE = Average Variance Extracted; Diagonal numbers in bold are square root of AVE

5.7 OVERALL DISCUSSION OF STUDY 2

The current chapter detailed the development and testing of the PSB-WD scale.

The development of the PSB-WD scale involved three phases: 1) item generation and

expert panel review, 2) piloting the survey, and 3) testing the scale’s dimensionality,

reliability and validity. First, the item generation phase pooled 84 items that were

reduced to 47 items through a group decision making technique conducted by the

research team. An expert panel was consulted to further give feedback on the items

and the item pool were further reduced to 42 items. Five items with low clarity and

low content validity were removed and six items were revised based on the feedback

by the expert panel. The initial PSB-WD scale with 42 items were categorised into six

dimensions: feedback inquiry, helping/volunteerism, problem prevention,

stewardship, voice and changing organisational policies and procedures.

The pilot testing of the 42 item PSB-WD scale revealed that, generally, most

items remained within the expected factors. For instance, the items for

volunteerism/helping, voice and feedback inquiry were originally written for these

factors. However, items written for some dimensions revealed mixed results,


particularly for the stewardship and problem prevention dimensions. For instance, it

was evident that problem prevention could be divided into fixing safety issues and

vehicle maintenance, while stewardship could be divided into protection, intervene

and knowledge-sharing.

Nevertheless, the results of the pilot study demonstrated the dimensionality of

the scale and assessment of the internal reliability of the extracted factors was

acceptable (all factors had a Cronbach α above 0.70). The PSB-WD scale was also

further reduced to 29 items, removing the items with poor communalities and high

cross-factor loadings.

The last phase of the scale development process involved an EFA to further

explore the factorability of the PSB-WD scale and confirmation of its factor structure

via CFA. The EFA demonstrated six factors: educate, feedback inquiry, voice,

intervene, fixing safety issues and vehicle maintenance (See Figure 5.3). The results

further clarified the factors found from the pilot studies, suggesting that vehicle

maintenance and fixing safety issues were indeed separate factors (which included

items originally written for the problem prevention dimension). Educate and intervene

were also separate factors (which included items originally written for the stewardship

and helping/ volunteering dimensions). Feedback inquiry and voice were the only

factors that stayed relatively stable between the three processes. From this process,

the psychometric properties of the scale demonstrated sound reliability and validity

(e.g., construct, convergent, divergent validities). The final version of the PSB-WD

scale had 21 items.

102Chapter 5: Development and Testing of Proactive Safety Behaviour Measure within the Work Driving Context

PSB-WD

Study 2b

Feedback Inquiry


Problem Prevention

Fixing safety issues

Vehicle Maintenance

Stewardship

Protection

Intervention

Knowledge Sharing Voice

PSB-WD

Study 2c

Feedback Inquiry

Educate

Fixing safety Issues

Vehicle Maintenance

Intervene

Voice

Figure 5.3 Hypothesised dimensions of proactive safety behaviours within the work driving context from the item generation (PSB-WD; left) and the dimensions extracted from PCA using the pilot study data (middle) and dimensions extracted from the EFA and CFA (left).

PSB-WD

Study 2a

Feedback Inquiry


Problem Prevention

Stewardship

Voice

Changing of organisations policies and procedures


5.7.1 Theoretical and Practical Implications

The purpose of the current study was to identify proactive safety behaviours that

are prevalent to the context of work driving and to develop a measure for these

behaviours. Adapting existing measures on proactive safety behaviours and proactivity

originally developed for general work settings, the PSB-WD scale provided a clearer

picture on the proactive safety behaviours that employees may undertake while driving

for work.

While this scale was developed specifically for the work driving context, the

current study provided further evidence that proactive safety behaviours is a multi-

dimensional construct, even when applied in a new context (i.e., work driving).

However, the PSB-WD scale also demonstrated possible new dimensions of proactive

safety behaviours, which included: feedback inquiry, fixing safety issues and vehicle

maintenance. These dimensions were specifically developed from the work proactivity

measures by Parker and colleagues (Parker & Collins, 2010; Parker & Wang; 2015) -

measures that previous studies on proactive safety behaviours have not examined.

Although these dimensions were specifically developed for work driving safety, it is

possible that these dimensions could also be applied within the general safety field.

In addition, previous studies that examined similar constructs to proactive safety

behaviours within the road safety research (e.g., traffic safety citizenship by Finley et

al. [2015]) only examined the passengers’ role in intervening when a driver engages

in a risky behaviour (e.g., using a mobile phone while driving). The current study has

contributed to the work driving literature by providing a larger set of behaviours that

employees may engage in when taking initiatives in work driving safety (e.g., asking

feedback about their driving; educating other drivers of safety policies and

procedures).

Furthermore, the newly developed PSB-WD scale could be used as a diagnostic

tool for work driving safety performance. Inspection of the internal reliability and

validity of the scale also indicated that the developed scale has sound psychometric

properties. This psychometrically sound scale could be used as a complementary tool

alongside current measures of work driving safety (e.g., occupational driving safety

behaviours, self-reporting of crashes, safety incidents, traffic violations and near

misses). More specifically, instead of just looking at lagging indicators (e.g., crashes


and traffic violations), organisations could use this scale to measure other indicators

of safety performance among work drivers.

The scale could also be used as an education tool to cultivate these behaviours

in organisations. The PSB-WD scale identified various behavioural indicators of

proactive safety within the work driving context. Companies could use the current

research findings to develop workshops on various ways that employees could be

proactive about safety while driving for work.

5.7.2 Limitations and Future Studies

There were a number of limitations that should be noted and directions for future

research. First, the low response rate from the pilot study represented a limitation to

the current study. It could be argued that the low response rate could reflect the time

constraints and the length of the survey. However, it was necessary to initially retain

all questions given that the studies were testing the stability and structure of a new

measure. Furthermore, this limitation was addressed in third phase of the study with a

larger sample size of work drivers.

In addition, further studies are needed to examine the test-retest reliability of the

new measure. While the current scale has shown excellent internal reliability and

validity (e.g., construct and content), further testing should be conducted to ensure its

stability over time. For instance, test-retest reliability analyses could be carried out

over several time-points to ensure the scale’s consistency. In addition, further evidence

of the external validity of the scale could be assessed based on supervisors’ rating of

work drivers’ proactivity levels towards safety issues and critical events. While this

testing could be resource intensive, it would likely offer further insights in regards to

assessing the quality of the scale.

Next, the study utilised self-report measures to assess the proactive safety

behaviours among work drivers. Although self-report measures are often used in traffic

research, these measures are subjected to participant response biases (Newnam et al.,

2011). For instance, the recruitment procedure of seeking voluntary participation for

the survey may have contributed to selection bias, whereby employees self-selecting

into completing the survey may be already engaging in proactive work driving safety

behaviours. Those who may not be performing such behaviours may be less likely to

volunteer for participation. Future studies could attempt to utilise other methods of


measuring proactive work driving safety behaviours to reduce these biases, such as

observation or diary studies.

5.7.3 Concluding Remarks

The current study is motivated by the significant issue of work-related road

crashes which result in fatalities and serious injuries. Furthermore, there is currently a

lack of evidence regarding the identification of leading indicators for safety

performance among the work driving population. The current study aimed to make a

practical contribution in the work driving safety field by developing a measurement

tool that could examine these behaviours in the workplace and by investigating how


managing risks while driving for work.

This study suggests that the newly developed PSB-WD scale is a

psychometrically sound scale for assessing proactive safety behaviours among work

drivers. At present, no research has captured the proactive safety behaviours prevalent

in the work driving population. Therefore, this new instrument has the potential to

further inform work driving safety research. For instance, research could use the newly

developed scale to develop better understanding of the relationships between the

antecedents and outcomes specific to this behaviour. In addition, the scale could be

used as a diagnostic tool to identify target behaviours and conditions for interventions.

Further discussion of this suggestion appears in the general discussion chapter

(Chapter 7).

106 Chapter 6: Model Testing of Proactive Safety Behaviours within the Work Driving Context (Study 3)

Chapter 6: Model Testing of Proactive Safety Behaviours within the Work Driving Context (Study 3)


Study 1 summarised and synthesized the contextual and individual factors that

influence proactive safety behaviours, using studies conducted within the general

occupational safety field. As shown from the findings of the meta-analysis conducted

within the current program of research (Chapter 4, section 4.4), employees’ proactive

safety behaviour is influenced by a wide range of contextual factors and cognitive

mechanisms. In conjunction with the findings from the meta-analysis and the current

literature on the work driving safety field, the current study proposes a new model for

proactive safety behaviours that could be applied within the work driving context (see

Figure 6.1). The proposed model is guided by the framework established by Curcuruto

and Griffin (2017) on safety proactivity and tested using a sample of work drivers (N

= 300).

Can-do Motivation: Perceived Control

Reason-to Motivation: Felt Responsibility

Proactive Future Orientation & Motivational States

PROACTIVE WORK DRIVING SAFETY BEHAVIOURS


CONTEXTUAL ANTECEDENTS PROXIMAL ANTECEDENTS SAFETY BEHAVIOUR

Perceived Work Environment

Safety Climate

Leader-Member Exchange

Future Orientation: Anticipation Focus

Figure 6.1 Research model

Chapter 6: Model Testing of Proactive Safety Behaviours within the Work Driving Context (Study 3) 107

This chapter presents the results of this model testing, which represents the third

and final study of the current program of research. This chapter also aims to answer

RQ3: “What are the individual and organisational factors that promote and inhibit

proactive safety behaviours in work drivers?”

It is also expected that a myriad of factors would influence the proactive safety

behaviours performed by work drivers. The current program of research only

specifically examined the potential influence of safety climate, leader-membership

exchange (LMX) as context antecedents of proactive safety behaviours among work

drivers. The current program of research also specifically examined future orientation

and motivational states as the proximal factors of these behaviours. The arguments for

choosing these factors are discussed in the following sections of the current chapter.


Safety Climate

As mentioned in Study 1 (meta-analytic review), researchers within the general

occupational field have often examined safety climate to investigate safety behaviours

in the workplace. To engage in proactive safety behaviours, employees must perceive

that such behaviours are valued and important within their work environment (Griffin

and Neal, 2000). The organisation’s climate often sets the expectations on which

behaviours are valued and rewarded. A positive safety climate sets the expectation that

safety is critical within the organisation. Therefore, positive safety climate encourage

workers to conduct safety behaviours and practices in the workplace because working

safely is valued within their organisation (Griffin & Neal, 2000).

An extensive body of literature exists on the role of safety climate in

occupational safety and, as demonstrated in Study 1, safety climate has a significant

effect on proactive safety behaviours. Safety climate has also been studied in the work

driving literature and seen as an integral factor in work drivers’ safe driving

performance (Newnam et al., 2011; Wills, Watson & Biggs, 2009). Safety climate

within the work driving literature has been defined as, “the mental framework or set

of perceptions that drivers hold about fleet safety policies and practices in their

organisations” (Husband, 2011, p. 17). Using work drivers from various Australian

organisations that operate vehicle fleets, Wills et al. (2009) found that employees’

perceptions of safety climate significantly predicted current driving behaviours and


future driving intentions over and above demographic characteristics (age and gender)

and driving exposure (annual mileage and weekly driving hours). Safety climate

explained the largest unique variance of the driving behaviours compared to other

factors (i.e., safety attitudes, perceived behavioural control and driving experience;

Wills et al., 2009). In Will et al.’s (2009) study, employees’ perception of the

importance and practicality of safety rules, how safety issues are communicated within

the organisation and management’s commitment to safety were found to be the most

influential safety climate elements on work-related driving behaviours (Wills et al.,

2009). However, safety climate within the work driving literature is usually examined

as a predictor of risky driving behaviours in work drivers. Safety climate as an

antecedent to proactive safety behaviours within the work driving context has not been

previously examined.

Furthermore, current research on safety climate within the work driving context

tend to focus on the factor structure of safety climate and which safety climate factor

is most important in predicting employees’ driving behaviours (e.g., Wills et al., 2006,

2009). However, recent studies by Huang and colleagues have started to examine the

perceptions of safety climate from different levels of the organisation (e.g., group

versus organisation)17 using different samples of work drivers of heavy and light

vehicles (Huang, Ho, Smith, & Chen, 2006; Huang, Lee, McFadden, Rineer, &

Robertson, 2017; Zohar, Huang, Lee, & Robertson, 2014; Huang et al., 2013). Huang

et al. (2013) developed a multi-level safety climate measure that could be used within

the work driving setting. Using their multi-level measure of safety climate, Huang et

al. (2013) demonstrated that higher levels of organisational and group safety climate

were associated with lower involvement in their work driver’s near miss events and

work-related injury and illness. Huang et al. (2013) found that group-level safety

climate has a stronger relationship with near-miss incident reporting compared to

organisational-level safety climate. Furthermore, the level of safety climate showed a

17 As mentioned in Chapter 4, Zohar and colleagues (Zohar, 2008; Zohar and Luria, 2005) proposed a multi-level framework for safety climate which differentiates between the organisation and group-level as employees’ perception may change depending on the referent group (i.e., top management or direct supervisors). In particular, safety climate from the organisation-level perspective may involve policies and procedures established by the company, while the top management ensures that these policies and procedures are implemented and safety is promoted. Group-level safety climate is more focused on the direct supervisory and workgroup safety practices. However, the top management’s priorities (organisational-level) may not always align with the priorities of the unit-level supervisors and the workgroup (group-level).


differential effect on the outcome variables. For instance, higher group safety climate

is a significant predictor of lower absenteeism due to illness, organisational safety

climate did not show a significant relationship with absenteeism. From these studies

and findings of Study 1, it is possible that different levels of safety climate

(organisational versus group) could also have a potential differential effect on work

drivers’ proactive safety behaviours.

Leader-Member Exchange.

In addition to safety climate, the impact of leaders is usually examined within

occupational safety. Leaders play a critical role in encouraging workers to perform

safety behaviours and in shaping the safety climate within an organisation (Hofmann

& Morgeson, 1999; Neal & Griffin, 2002). Various leadership theories exist in the

literature, however, the current chapter focuses on the social exchange relationship

between leaders and subordinates and its influence on proactive work driving

behaviour. Specifically, the current study used the leader-member exchange (LMX)

theory for the proposed research model.

The LMX theory describes the quality of the relationship between leaders and

their followers. This theory offers a unique perspective on leadership due to its focus

on the ‘dyadic relationship’ between the leader and the subordinate (Graen & Uhl-

Bien, 1995). High quality LMX is characterised by trust, mutual respect, fairness and

support between leaders and their followers (Graen & Uhl-Bien, 1995). Within the

safety literature, research shows that having a high quality LMX is associated with

more open communication and increased value congruence (Hofmann & Morgeson,

2004). As a result, workers are encouraged to raise safety concerns and become more

committed to safety in their workplace (Hofmann & Morgeson, 2004). Hofmann and

Morgeson (2004) also found that employees with high quality relationships with their

leaders are more likely to perceive safety as a job responsibility and more likely to

engage in more safety citizenship behaviours. Didla et al. (2009) suggested that the

concept of safety citizenship behaviour is based on the principle of reciprocity and the

social exchange between members and leaders. If the reciprocal supervisor-employee

relationship is built upon mutual trust, support and fairness, workers will reciprocate

positively to his/her leader and engage in behaviours that they perceive their leaders

find important (Didla et al., 2009). Therefore, leaders who place a strong emphasis on

safety in the workplace are more likely to encourage safety behaviours in their workers


if they have a high quality relationship with their subordinates (Didla et al., 2009).

Hoffmeister et al. (2014) also argued that when leaders consider their employees

individually, employees are more open to generating ideas and solutions to safety-

related problems (i.e., active promotion of safety practices).

The results of Study 1 demonstrated that transformational leadership showed a

significant and moderate relationship with proactive safety behaviours. While

transformational and LMX taps into different aspects of leadership, with the former

describing leadership traits while the latter explaining the social exchange relationship

between leaders and subordinates, Clarke (2013b) argued that transformational

leadership also encompasses aspects of LMX. In particular, Clarke (2013b) suggested

that transformational leaders demonstrate individualised consideration, whereby a

leader is said to show interest in subordinates’ personal and professional development

and listens to their needs and concerns (which is an aspect of LMX).

Furthermore, research within the work driving field has previously examined the

role of LMX on safety driving behaviours and found that the quality of leader and

member exchange is a significant factor of safety driving performance (Newnam et al.,

2012). Newnam et al. (2012) argued that “participation in safe driving practices is

likely to be promoted through reciprocation of those behaviours valued by the

supervisor” (p. 32). Using a sample of healthcare employees who drive for work,

Newnam et al. (2012) studied the role of LMX on the safety information exchange

between supervisors and work drivers and driving performance. Safety information

exchange measures the frequency of discussions on driver safety with direct

supervisors within a two-week period. Their study found that high quality relationship

with leaders is associated with more frequent exchange of safety information and safer

driving performance (Newnam et al., 2012). In other words, LMX not only promoted

safety communication between supervisors and work drivers, but it also promoted

engagement in safe driving behaviours.

Therefore, instead of examining the influence of leadership traits on proactive

safety behaviours, LMX was used in the model as an antecedent to work drivers’

proactive safety behaviours. The current program of research proposed that the

relationship between the leaders and subordinates would have an impact on work

drivers’ engagement in proactive safety behaviours.


6.1.2 Proximal Antecedents

Future orientation and motivational states are the central aspects of Curcuruto

and Griffin’s (2017) model of safety proactivity. As discussed in Chapter 4, Curucuto

and Griffin (2017) argued that several cognitive-mechanisms may impact directly on

one’s motivation to act in a proactive manner, capability and commitment drivers and

future orientation. The capability and commitment mechanisms were based on the

expectancy-valence theory of motivation, while future orientation states were based

on the regulatory focus theory. Since these concepts were already discussed in Chapter

4, only a brief overview and their relation to the research model are provided in the

following sections.

Proactive Future Orientation – Anticipation-Focus

Recently, Curcuruto et al. (2016) proposed the framework of future orientation

and its relationship with safety management. Curcuruto et al. (2016) argued that

individuals who have the propensity to anticipate risks and those who strive to improve

safety in the workplace beyond the minimal standards are more likely to behave

proactively for safety. These individuals have the future orientation of ‘anticipation’

or ‘improvement’.

The construct of future orientation taps into the promotion-focused concept

within the self-regulation focus theory (Curcuruto et al., 2016). The self-regulation

focus theory describes the two cognitive approaches that individuals engage in when

deciding which behaviours to perform during a goal pursuit (Higgins, 1997; Henning

et al., 2009). These two cognitive approaches are promotion-focused or prevention-

focused (Higgins, 1997). Curcuruto et al. (2016) applied this concept of promotion-

focus in occupational safety contexts. They suggested that individuals who are

promotion-focused are more likely to anticipate risks and to continuously improve

safety within the workplace (Curcuruto et al., 2016).

Anticipation-focused and improvement-focused individuals are more likely to

suggest new or different ways to do things more safely (Curcuruto et al., 2016). These

individuals are also more willing to think about ways to improve safety at work even

if activities are running smoothly and there is no evidence of apparent threat

(Curcuruto et al., 2016). Using a sample of chemical and manufacturing operators,

Curcuruto et al. (2016) found that individuals who anticipated risks and who had the

propensity to improve the safety of their work environment were more likely to engage


in proactive safety behaviours. More specifically, individuals with anticipation

orientation and improvement orientation were more likely to voice safety concerns and

suggest safety improvements, and were less likely to engage in risk-taking behaviours

The concepts of anticipation and future orientation have not yet been applied in

the work driving context, however, similar constructs had been studied in the general

road safety research, particularly with risk anticipation (e.g., McKenna, Horswill, &

Alexander, 2006; Kinnear, Kelly, Stradling, Thompson; 2013). For instance, McKenna

et al. (2006) found that training individuals in anticipating hazards on the road were

associated with reduced risk driving. They argued that risk taking behaviour, while

deliberate, could also reflect failure to appreciate the danger within the immediate

environment. In line with this argument, anticipation focused individuals may pre-

empt risks and hazards when driving for work. Research on improvement orientation

within driving safety is still unclear. Therefore, the current study focused on the role

of anticipation orientation on proactive safety behaviours within the context of work

driving.

Proactive Motivational States – Can Do and Reason To

While safety climate and high-quality relationship with leaders create an

environment that encourages individuals to engage in proactive safety behaviours,

individuals must actually be motivated to perform these behaviours (Andriessen, 1978;

Griffin and Neal, 2000). Individuals’ motivation to proactively engage in safety

practices is central to Curcuruto and Griffin (2017) model of safety proactivity.

Curcuruto and Griffin (2017) argued that motivation directly influence proactive safety

behaviours and act as proximal antecedents for such behaviours.

As mentioned in Chapter 4, Curcuruto and Griffin (2017) argued that several

cognitive-mechanisms may directly impact on one’s motivation to act in a proactive

manner. These motivational states were referred to as: capability and commitment

drivers. These motivational states were based on Vroom’s (1964) expectancy-valence

theory. In its simplest form, the expectancy-valence theory proposes that an individual

carries out a behaviour depending on (a) the perceived value of the outcome (valence)

and (b) the perceived probability of achieving that outcome (expectancy; Pinder,

2008). For instance, a worker will engage in a work behaviour (e.g., staying late at

work) if they believe that this behaviour is likely to result in valued rewards and

outcomes (e.g., getting a promotion; Pinder, 2008).


In line with the expectancy-valence theory, Curcuruto and Griffin (2017)

conceptualised the capability drivers as the ‘can-do’ motivational processes which

relates to one’s self-efficacy and belief that they can actually carry out the behaviour

(similar to the expectancy). On the other hand, commitment drivers refer to the

‘reasons’ why individuals should carry out the proactive behaviour (similar to valence

and instrumentality). Within these two broad categories of ‘can-do’ and ‘reason-to’

motivational drivers, Curcuruto and Griffin (2017) argued that there are 4 dimensions

of motivational states. The capability ‘can-do’ drivers include the components of:

1) role-breadth self-efficacy – which describes the extent in which an individual

is confident that they are able to carry out proactive safety activities and

perform safety initiatives, and

2) perceived control – which describes the extent in which individuals believe

they have “control” or are influential in shaping the safety systems and

procedures within their organisation.

On the other hand, the commitment ‘reason-to’ drivers include:

1) psychological ownership – which refers to the feeling of possession to an

‘object’ (material or immaterial) within an organisation. Within the safety literature,

psychological ownership of safety relates to the individual’s belief that their safety at

work is ‘their own’, and

2) felt responsibility – which refers to one’s perceived responsibility of safety

within the workplace.

While these dimensions of motivational states have been examined in the general

proactivity literature (e.g., Bindl, 2010; Parker et al., 2010; Parker et al., 2006), these

concepts have not been fully examined in the general occupational safety research and

especially within the work driving context. Available evidence suggests that these

motivational states were positively correlated with safety initiative, safety voice and

safety stewardship and with negative relationships with risk-taking behaviours

(Curcuruto et al., 2016). The meta-analysis conducted as part of this program of

research also showed that self-efficacy and perceived control were significant

proximal indicators of proactive safety behaviours, with perceived control showing a

stronger effect (rcor = 0.75) compared to self-efficacy (rcor = 0.42).


Although these motivational states have not been extensively examined within

the work driving context, similar concepts have been studied. For example, Banks,

Davey, and Biggs (2010) studied the perceived shared responsibility with road safety

outcomes using a sample of work drivers in Australia. They found that work drivers

who felt that responsibility in achieving work-related road safety is shared within the

organisation, were more likely to engage in safe driving behaviours. Specifically, these

individuals were less likely to drive while fatigued or to drive while multi-tasking

(Banks et al., 2010).

Perceived control had also been examined within the general road safety

research, but usually as an aspect of the Theory of Planned Behaviour, rather than as

a motivational state. In Otto et al.’s (2016) research on traffic safety citizenship, they

found that individuals who felt more confident and comfortable in speaking up were

more likely to intervene when a driver is conducting a risky behaviour (e.g., using their

phone while driving or not wearing a seatbelt).

The current study specifically studied the effect of felt responsibility and

perceived control on work drivers’ proactive safety behaviours. In addition, an indirect

effect between safety climate and proactive safety behaviours, via anticipation focus,

perceived control and felt responsibility, was proposed as previous studies have

demonstrated that safety climate has a direct and indirect influence on safety

behaviours via safety motivation and regulatory focus (Wallace et al., 2016; Christian

et al., 2009).

6.2 NON-PROACTIVE ANTECEDENT, MEDIATORS AND SAFETY COMPLIANCE

Although proactive safety behaviour is the focus of the current research, a

discussion of safety compliance is required so as to have a comprehensive

understanding of proactive safety behaviours. Violation of rules and procedures is

often seen as a major cause of workplace accidents and injuries (Zohar, 2008).

Therefore, complying with safety policies and procedures are integral in the

functioning of organisations and ensuring the safety and wellbeing of everyone at

work. As mentioned earlier, safety compliance (along with safety participation) is an

important dimension of Griffin and Neal's (2000) safety performance framework.

Safety compliance consists of behaviours synonymous with adhering to rules,

regulations, policies, and procedures relating to safety many of which are typical in


traditional safety strategies (e.g., wearing personal protective equipment and

conducting risk assessment strategies according to policies and regulations). Safety

compliant behaviours are essential in the work driving context because of the inherent

risks associated with driving for work. For example, various regulatory procedures are

in place to manage work driving fatigue (e.g., limiting hours of driving, providing

minimum hours of rest; Gander et al., 2011). These regulations ensure the safety of the

driver and the community and failure to comply with these regulations could have

devastating consequences in terms of contributing to road crashes.

Indeed, safety compliance is vital and often seen as an important prerequisite for

safe commercial vehicle operations. However, simply focusing on these behaviours as

a measure of safety performance is limited and unlikely to be sustainable in the long

run (Wishart et al., 2019). If the safety operation of an organisation simply relies on

workers complying with rules and regulations, engagement with safety would be

minimal and would not encourage desire for improvement (Curcuruto et al., 2015). In

routine and predictable work, complying with safety rules and regulations could be

adequate for maintaining safe practices (Wishart et al., 2019). However, if work has

less routine and predictability, such as driving for work (due to various road

conditions, weather, other drivers on the road, etc.), complying with safety procedures

may not be sufficient (Zohar, 2008). Proactive safety behaviours may be necessary to

guide safe practices where compliance to rules and procedures may fall short (Zohar,

2008). By including proactive safety behaviours in addition to safety compliance as

key safety performance indicators in the workplace, organisations are providing a more

comprehensive management of safety in routine work practices as well as ensuring

workers’ safety in less predictable circumstances such as a sudden adverse weather

conditions while driving or noticing a vehicle defect during a delivery (Zohar, 2008).

Furthermore, since the current program of research used a recently developed

scale to measure proactive work driving safety behaviours, safety compliance was

added to the research model to test for discriminant validity. While proactive safety

behaviours and safety compliance are closely related (Christian et al. 2009), different

mechanisms are involved in these two constructs. For instance, Griffin and Hu (2013)

found that leadership styles differ on their impact on employees’ safety compliance

and safety participation behaviours. Leaders who monitored workers’ mistakes and

safety violations were more likely to promote safety compliant behaviours among their


subordinates. On the other hand, leaders who encouraged members to learn from

mistakes and place high value on team members’ safety were more likely to promote

safety participation among employees. It could be argued that different situational,

individual and cognitive mechanisms are involved in safety compliance and proactive

safety behaviours.

Therefore, to examine the discriminant validity of the model and to provide a

comprehensive account of safety performance in work drivers, the current research

also examined the different mechanisms involved in safety compliant behaviours along

with proactive safety behaviours. From the literature, it was expected that safety

climate and the quality of the relationship between leaders and members will be

associated with safety compliance, but the strength of the relationship may be weaker

(as was demonstrated by Christian et al.’s [2009] meta-analysis). Furthermore, the

cognitive mechanisms involved in engaging with safety compliant behaviours were

also expected to be different from cognitive mechanisms involved with proactive

safety behaviours. For instance, instead of having an anticipation focus orientation,

safety compliance may be promoted by a preventative orientation. Individuals with

preventative orientation may comply with the rules and regulations to avoid being

penalised when caught and may comply with rules and regulations are fear of negative

repercussions instead of feeling responsible for their safety and the safety of others

(Aryee & Hsiung, 2016; Curcuruto & Griffin, 2017).


6.3 STUDY HYPOTHESES

The following hypotheses were developed from the aforementioned evidence

base model of the current program of research (See Figure 6.2):

Direct Effects. Having a positive organisational and group safety climate (H1)

and high quality leadership-member relationship (H2) will have a direct positive

relationship with work drivers’ proactive safety behaviours. Anticipation focus,

perceived control and felt-responsibility will also have a positive relationship with

proactive work driving safety behaviours (H3).

Indirect Effects. Future orientation and motivational states will act as mediating

variables between the safety climate and proactive safety behaviours. In other words,

safety climate (group and organisational level) will also have a positive relationship

with work drivers’ proactive safety behaviours via anticipation focus, perceived

control and felt responsibility (H4).

Non-proactive orientation, motivation and outcomes. Although not a core

focus of the program of research, it is important to demonstrate that the concept of

proactive safety behaviours is different from safety compliance. Therefore, an

additional outcome variable, safety compliance, along with its proposed antecedent

Figure 6.2 Research model with hypotheses


variable, preventative focus was added to the model. Therefore, it was hypothesised

that preventative focus and safety compliance will show a differential effect towards

the proposed model of proactive safety behaviour. Specifically, it was expected that

preventative focus will have a positive relationship with safety compliance but will

have a negative relationship with proactive safety behaviours. It was also expected that

safety climate will have a direct relationship, and an indirect relationship, with safety

compliance via preventative focus. LMX will also have a direct positive relationship

with safety compliance.

6.4 METHOD

The data used for Study 3 was the same data collected Study 2c, therefore, the

methodology, data cleaning and participant demographics and work characteristics for

Study 3 can be found in the Chapter 5. However, the measures that relate to the data

analysis of the current chapter were discussed in the following section. See Appendix

E and F for the participant information sheet and survey used for Study 3.

6.5 MEASURES

Participants were asked to indicate their demographics (i.e., age and gender),

years of having had a driving licence and work-driving related activities such as:

number of hours driven to work per week; kilometres driven to work annually; amount

of time worked in the organisation; type of vehicle driven for work; and, typical type

of road driven for work. Participants were also questioned on how many crashes and

near-misses they had experienced and traffic offences they had received while driving

for work in the past year. The following measures were used in the survey:

6.5.1 Proactive Safety Behaviours within the Work Driving Context (PSB-WD).

Proactive safety behaviours were assessed using the newly developed 21-item

proactive safety behaviour scale within the work driving context (PSB-WD). The

results from the analysis underpinning this measure are presented in Study 2c and

revealed a six-factor scale of PSB-WD: Education, Voice, Fixing Safety Issues,

Vehicle Maintenance, Feedback Inquiry and Intervene. Participants rated the extent to

which they performed these behaviours on a 5-point Likert-type scale of 1 (never) to

5 (always).


6.5.2 Safety Climate Survey – Short Form.

A scale of safety climate specifically developed for work drivers by Huang et al.

(2013) was used in this study. This scale measures safety climate at the organisational-

and group-levels within the work driving context. Zohar (2008) argued that perception

of safety climate changes depending on the referent group (i.e., organisational level

versus group-level). Organisation-level safety climate refers to the workers’

perception of the safety procedures and policies established by the company and top

management (and was assessed by 6 items). Group-level safety climate describes the

perceptions that workers have of their direct supervisor and workgroup safety practices

(and was assessed by 6 items). Participants rated the extent they agreed with each

statement on a 5-point Likert-type scale of 1 (strongly disagree) to 5 (strongly agree).

Higher scores indicate a more positive perception of organisational and group safety

climate.

6.5.3 Leader-Member Exchange (LMX).

The Leader-Member exchange (LMX) scale developed by Graen and Uhl-Bien

(1995) was used to measure the quality of the relationship between work drivers and

their supervisors. LMX was measured using seven items (e.g., “My supervisor

understands my job problems and needs” / “I understand the problems and needs of

my work drivers”). Participants rated the extent they agreed with each statement on a

5-point Likert-type scale of 1 (strongly disagree) to 5 (strongly agree). Higher scores

indicated a more positive exchange relationship. Previous research showed good

internal consistency and validity scores for the scale and this measure had been

previously applied within the general safety research (Hofmann et al., 2003) and work

driving context (Newnam et al., 2012).

6.5.4 Anticipation Focus, Perceived Control and Felt Responsibility.

Anticipation focus, perceived control and felt responsibility were measured

using the proactive safety (PRO-SAFE) scale developed by Curcuruto et al. (2016).

The PRO-SAFE scale measures the cognitive and psychological mechanisms (i.e.,

future orientation and motivation) of proactive management of safety and incident

prevention in the workplace. For the current study, the dimensions of anticipation

focus, perceived control and felt responsibility were utilised. Previous research has


demonstrated solid psychometrics for the PRO-SAFE scale using samples from high

risk industries (e.g., chemical and manufacturing; Curcuruto et al., 2016).

The scale was adapted to suit the work driving context, therefore confirmatory

factor analysis (CFA) was conducted to assess the validity of the adapted scale. The

model showed an acceptable fit, χ2(51) = 170.06, p < .001; CMIN/DF = 3.34, CFI =

0.94, TLI = 0.92, NFI = 0.91, RMSEA = 0.09, SRMR = .09). All items demonstrated

significant loadings to the hypothesised factors; however, some had low factor

loadings (less than 0.06). These items were deleted (n = 2) and the second model

showed a better fit, χ2(32) = 76.50, p < .001; CMIN/DF = 2.39, CFI = 0.97, TLI = 0.96,

NFI = 0.96; RMSEA = 0.07; SRMR = 0.06). The final version of the scale used with

the current analysis had 4 items measuring anticipation focus (e.g., “I anticipate risks

or safety problems when driving for work, thinking of the possible alternative

scenarios if problem arises”), 3 items measuring perceived control (e.g., “I recognise

that I am able to make significant contributions to the work driving safety within my

department”) and 3 items measuring felt responsibility (e.g., “I feel that it's my

responsibility to discuss work driving safety issues with my co-workers”). Participants

rated their agreement with each statement on a 5-point Likert-type scale of 1 (strongly

disagree) to 5 (strongly agree). Higher scores indicate higher levels of proactive safety

motivation and future safety orientation.

6.5.5 Prevention Focus and Safety Compliance.

The measure for prevention orientation that was utilised in the current study was

adapted from several existing measures on prevention focus (Wallace & Chen, 2006;

Neubert, Kacmar, Carlson, Chonko, & Roberts, 2008). In addition, the safety

compliance measure used for the current study was adapted from Neal and Griffin’s

(2006) study. Both measures were adapted to suit the work driving context. Four items

were utilised for safety compliance (e.g., “I follow the work driving safety policies and

procedures”) and three items for prevention focus (e.g., “I try to avoid making mistakes

or violating rules when driving for work because I do not want to get in trouble”). The

items for prevention focus and safety compliance were subjected to a CFA due to the

applied changes. The results of the CFA a good model fit, χ2(6) = 9.78 p < .001;

CMIN/DF = 1.63, CFI = 0.96, TLI = 0.99, NFI = 0.99, RMSEA = 0.47; SRMR = .03).

For preventative focus, participants rated the extent they agreed to each statement on

a 5-point Likert-type scale of 1 (strongly disagree) to 5 (strongly agree). For safety


compliance, participants rated the extent they performed these behaviours on a 5-point

Likert-type scale of 1 (never) to 5 (always). Higher scores indicate higher levels of

preventative focus and safety compliance.

6.6 DATA ANALYSIS

6.6.1 Overview of Data Analysis

Data was cleaned and relevant assumptions for Structural Equation Modelling

(SEM) were checked. To assess interrelationships between the variables and the

complete structural model, SEM was used for the analysis. The current data analysis

followed the two-step modelling approach (Anderson & Gerbing, 1988). Using the

two-step modelling approach, the measurement model was initially assessed using

exploratory factor analysis (EFA) and CFA, where the measurement model is adjusted

and fitted to the data prior to testing the structural model (see Chapter 5). Once the

developed scale had been assessed, the testing of the structural model was conducted

– where the structural relationship between the measured variables and latent

constructs was analysed. SEM is a powerful multivariate analysis technique that has

the capacity to estimate multiple and interrelated relationship within a single analysis,

while addressing the issues of measure-specific errors (Hair et al., 2010). Therefore,

this statistical technique was chosen over regression mediation modelling. For the

current analysis presented herein, the potential causal relationship between the

variables were assessed using a latent mediation model via path diagrams with

Maximum Likelihood estimation method to fit the model. Regression estimates and

modification indices were examined to improve the model fit if necessary.

6.7 RESULTS

6.7.1 Treatment of Missing Data

As mentioned in Study 2c (see Chapter 5), there was minimal missing

information across variables and data were missing completely at random, as

demonstrated by a non-significant Little’s MCAR test. To preserve power, subscale

mean imputation was used to deal with the missing values on the variables of interest.

Missing data was replaced by calculating the participant’s mean score on each

subscale. Subscale mean imputation is a reliable method for preserving power, when

missing data is minimal, which is the case in this study (Schafer & Graham, 2002).

Schafer and Graham (2002) demonstrated that this method does not create bias in


preliminary data. Thus, the mean substitution should not have impacted the main

analyses of the current study. There were variables that did not have enough

information (e.g., no response or only 1 response to a variable item) and, therefore,

mean imputation was not possible. Since bootstrapping and use of modification indices

for SEM require full data information, cases with missing data, where mean imputation

was not possible, were deleted (n = 6).

6.7.2 Aggregation of Items

Although several authors have suggested that N = 200 is a sufficient sample size

to achieve statistical power when conducting a SEM analysis, it is acknowledged that

the sample size is still restricted in terms of testing a full measurement and structural

model due to the large number of variable items (e.g., Little, Cunningham, Shahar, &

Widaman, 2002; Matsunaga, 2008). As such, the measured observed variables were

aggregated to form a mean for each latent construct. The structural path analysis was

conducted using the aggregated scores. While this reduces the complexity of the model

(Little et al., 2002), it is still possible to assess the relationships among the variables.

Using the aggregated scores as indicators of the latent construct, a power analysis was

carried out to ensure that the current sample size (N = 294) was of sufficient statistical

power to detect significant effects. Using the G*Power program, computation of

achieved power was 0.99, revealing sufficient power.

6.7.3 Assumption Checking

Following the assessment and treatment of missing data, the scales were assessed

for non-normal distribution of data as SEM requires variables to be normally

distributed (Hair et al., 1995; Kline, 2000). Therefore, prior to the hypothesis testing,

the raw data were examined for univariate outliers and assumptions of normality.

Visual inspection of histograms and box-plots revealed univariate outliers on some of

the variables (i.e., PSB-WD feedback, vehicle maintenance, preventative focus,

anticipation focus, perceived control and felt responsibility). However, no extreme

outliers were observed in any of these variables (none with a z score of more than 3.29;

Field, 2013). Kolmogorov-Smirnov test of normality was examined and the results

revealed non-normality across all variables. Inspection of the histograms and statistics

for skewness and kurtosis also confirmed breaches of normality. When assumptions

of residual’s normality and homoscedasticity are not met, the standard errors are

usually biased, which could inflate the confidence intervals of the regression


coefficients (Field, 2013). Therefore, due to these violations in normality, bootstrap

analysis of 10,000 samples was carried out on all analysis. Bootstrap analysis does not

rely on the traditional central limit theorem and its assumptions (Efron & Tibshirani,

1986; Field, 2013).

Checks were also conducted to determine whether any differences on proactive

behaviours existed depending on the where the data was collected (online survey

versus hard-copy; See Appendix G). A series of independent t-tests were conducted

comparing the differences between PSB-WD scores from the online and hardcopy

collection methods. Significant differences were only found in regards to vehicle

maintenance, t(238.76) = 3.67, p = .001, BCa 95% CI [-1.89, 0.55], feedback inquiry

scores, t(191.52) = 2.98, p = .003, BCa 95% CI [0.33, 1.64], and intervene, t(289.00)

= -1.91, p = .037, BCa 95% CI [-1.51, -1.04] and the effect sizes were small (i.e.,

Cohen’s d = 0.29, Cohen’s d = 0.33, and Cohen’s d = 0.29, respectively). The sample

who completed the survey online were less likely to report vehicle maintenance and to

intervene but more likely to ask for feedback about their driving. The other proactive

safety behaviours did not show a significant difference between the data collection

method. Due to the small effect sizes, it is assumed that the data collection method had

no substantial influences on the results. Furthermore, using all of the sample ensured

the statistical power of the analysis and, therefore, the overall sample was used.

Common Method Variance Test

Due to the cross-sectional nature of the data, common method bias (CMB) is a

potential problem (Lindell & Whitney, 2001). CMB generally refers to the situation

when the correlations between constructs are due to the method used not the constructs

themselves (Podsakoff, Mackenzie, Lee, & Podsakoff, 2003). To examine if CMB was

a threat in this study, Harman's single-factor test was used (Podsakoff, et al., 2003).

An EFA was conducted with the number of factors fixed to one, using the items

retained from the CFA. Inspection of the unrotated factor solution showed that the first

factor only accounted for 29.1% of the total variance. CMB is considered to exist if

one factor explains more than half of the variance in all the items. Using this test, CMB

was not considered a problem within the study.


6.7.4 Description of Data

The descriptive statistics and reliabilities of the subscales are presented in Table

6.1. All subscales were associated with reliability scores of Cronbach’s α above 0.70,

thus indicating internal consistency was acceptable (Kline, 2000).

Examining the patterns of proactive safety behaviours, Table 6.1 demonstrates

that the sample were more likely to report vehicle maintenance behaviours as indicated

by the higher average scores and negative skew compared to the other proactive safety

behaviours. Participants were also less likely to engage in feedback inquiry as

demonstrated by the low average scores and positive skew.

As shown in Table 6.1, most of the variables were positively and significantly

correlated with proactive safety behaviours. Generally, the safety climate variables

showed a range of positive and moderate correlations between the proactive safety

behaviour variables, with organisational safety climate having the strongest positive

correlation with vehicle maintenance (r = .35), and group safety climate having the

strongest positive correlation with educate (r = .38). The correlations between LMX

and the proactive safety behaviour variables ranged between low to moderate (r = .18

to r = .29).

Anticipation focus was shown to have moderate strength correlations with the

proactive safety behaviours. Perceived control and felt responsibility also

demonstrated moderate to high correlations with proactive safety behaviours. The

highest correlation observed was between educate and felt responsibility, r = .52.

Compared with the work environment variables, the proximal variables were

generally more highly correlated with proactive safety behaviours. This difference in

correlation strength could indicate the temporal ordering assumption that is required

in mediation analysis. Within mediation analysis, it is assumed that the mediating

variables (i.e. focus and motivation variables) are more highly related to the outcome

variable (i.e., proactive safety behaviours). Therefore, temporal ordering can be

assumed.


Table 6.1 Means, Standard Deviations, Cronbach Alpha and Inter-correlations between the variables

PSB-WD Variables

M SD

Skew α SC-Org

SC-Group LMX

Anticipation Focus

Perceived

Control

Felt Responsibility

Prevention

Focus

Safety Compliance Educate Voice

Fixing Issues VM

Feedback

Intervene

PSB-WD Total

SC-Org 20.34 6.00 -.30 0.93 -

SC-Group 17.74 6.55 -.11 0.94 .74** -

LMX 27.51 5.74 -.92 0.94 .44** .41** -

Anticipation Focus

15.76 2.68 -.78 0.87 .26** .27** 0.10 -

Perceived Control 10.65 2.44 -.45 0.84 .43** .49** .27** .39** -

Felt Responsibility

9.99 2.64 -.45 0.87 .32** .41** .25** .31** .61** -

Prevention Focus 7.53 1.81 -.66 0.81 .14* 0.05 0.02 .20** 0.06 0.05 -

Safety Compliance

16.39 4.04 -1.20 0.83 .35** .24** .28** .24** .25** .25** .17** -

PSB-WD Variables

Educate 14.08 5.63 .06 0.90 .29** .38** .21** .40** .48** .52** 0.01 .36** -

Voice 9.01 3.56 .01 0.92 .20** .23** .22** .39** .45** .32** -0.01 .35** .65** -

Fixing Issues

13.50 4.24 -.38 0.87 .26** .21** .22** .40** .40** .38** -0.01 .42** .61** .56** -

Vehicle Maintenance

12.19 3.30 -1.25 0.84 .35** .18** .29** .32** .36** .27** 0.10 .47** .40** .42** .51** -

Feedback Inquiry 5.50 3.08 1.33 0.90 .24** .36** .18** .28** .31** .36** 0.10 .19** .56** .37** .33** .16** -

Intervene 10.86 3.05 -.58 0.86 .28** .25** .24** .34** .42** .41** 0.04 .48** .48** .42** .46** .39** .24** -

PSB-WD Total 65.14 17.08 -.30 - .36** .37** .30** .48** .55** .52** 0.06 .51** .88** .78** .80** .64** .60** .66** -

Notes. N = 294; ** Correlation is significant at the 0.01 level (2-tailed); * Correlation is significant at the 0.05 level (2-tailed); M = Mean; SD = Standard Deviation; α = Cronbach’s alpha; SC-Org = Organisational Safety Climate; SC-Group = Group Safety Climate; LMX = Leader-member exchange; PSB-WD = Proactive Safety Behaviours within the Work Driving Context; VM = Vehicle Maintenance; Highlighted cells demonstrate significant bivariate inter-correlations between the antecedents and PSB-WD variables


6.7.5 Who are more likely to engage in Proactive Safety Behaviours?

Before going on to the hypothesis testing, several tests were conducted to assess

demographic and work variables and its potential relationships with proactive safety

behaviours. The relationships between the continuous and ordinal variables (i.e., hours

of driven to work each week, annual kilometres of driving to work, employed time in

the current organisation, crashes in the past year, near misses in the past year and traffic

offences in the past year) and the proactive safety behaviour variables were carried out

using bivariate correlation analyses. For the categorical variables (e.g., gender, work

vehicle used, work road), a series of independent t-tests (gender and work road) and

One-way ANOVAs were conducted (vehicle used).

As seen in Table 6.2, hours per week, annual kilometres, employed years were

significantly correlated with most of the proactive safety behaviours. Generally, the

hours driven per week and annual kilometres driven to work had positive and

significant relationships with most of the proactive safety behaviours, except for

intervene and voice. Furthermore, positive and significant relationships were found

between years of employment with voice, fixing safety issues, vehicle maintenance

and intervene.

Only a small number of participants who self-reported to having experienced

these events in the past year. More specifically, only n = 39 (13.7%) has reported to

have experienced a crash and n = 50 (17.0%) has reported committing traffic

violations. More participants reported having experienced a near-miss n = 183

(62.2%). Self-reported crashes, near misses and traffic offences in the past year did not

demonstrate any significant relationships with proactive safety behaviours, except for

a significant negative correlation between vehicle maintenance and self-reported

traffic offences.


Table 6.2 Bivariate correlations between Proactive Safety Behaviour variables and work driving exposure and risk

Hours per

week Annual

Kilometers Employed

years Crashes in

the past year

Near misses in the past

year

Traffic offences in

the past year

Educate .20** .18** 0.10 0.02 0.07 0.02

Voice .22** 0.11 .18** 0.04 -0.01 -0.08

Fixing Safety Issues .19** .18** .13* 0.08 -0.07 0.02 Vehicle Maintenance .17** .14* .19** 0.05 -0.03 -.13*

Feedback Inquiry .13* .13* 0.03 0.10 0.06 0.03

Intervene 0.04 -0.01 .15** -0.10 -0.11 -0.05 Notes. N = 294; ** Correlation is significant at the 0.01 level (2-tailed); * Correlation is significant at the 0.05 level (2-tailed); PSB-WD = Proactive Safety Behaviours within the Work Driving Context; Crashes, near misses and traffic offences are self-reported.

When assessing the potential influence of gender on proactive safety

behaviours, the only significant difference found was on feedback inquiry (see

Appendix H). Males were more likely to ask for feedback about their work driving

compared to females, t(279.41) = 2.43, p = .011, 95% BCa 95% CI [0.11, 1.51]. No

significant differences were found between male and female in their likelihood of

performing other proactive safety behaviours.

Furthermore, differences in proactive safety behaviours were compared

between employees who have a supervisory role compared to those with no

supervisory role (Appendix I). Significant differences between these two groups were

significant for all types of proactive safety behaviours. In general, participants with a

supervisory role were more likely to report engaging in proactive safety behaviours

compared to those without a supervisory role.

Differences on the type of road that the current sample usually drive on were

also considered. Due to the small sample size of drivers who reported to primarily

drive on dirt roads (n = 3), the independent samples t-tests were only conducted

between work drivers who reported driving on asphalt or bitumen roads compared to

a combination of asphalt and dirt roads. Drivers who reported driving on the

combination of asphalt and bitumen roads were more likely to perform most of the

proactive safety behaviours compared to those who drive on asphalt or bitumen roads

only (See Appendix J). The only exception to this finding was that, there were no

significant group differences found on the intervene variable.

Differences in proactive safety behaviour performance as a function of vehicle

type were also assessed. Overall group differences were only found for voice, F(3,


291) = 8.79, p =.001, and vehicle maintenance, F (3, 291) = 8.79, p =.001. Games-

Howell Post-hoc analysis was used due to the unequal sample sizes and breaches in

normality assumptions. The results showed that those who drive heavy vehicles (Mdiff =

2.29, BCa 95% CI [0.84, 3.75]) and SUV or 4WD (Mdiff = 1.52, BCa 95% CI [0.44,

2.59]) were more likely to voice safety issues compared to those who reported driving

a smaller car. Similarly, those who drive heavy vehicles (Mdiff = 1.74, BCa 95% CI

[0.22, 3.25]) and SUV or 4WD (Mdiff = 1.53, BCa 95% CI [0.64, 2.41]) were more

likely to perform vehicle maintenance compared to those who drive a smaller car.

6.7.6 Hypothesis Testing using Structural Equation Modelling

Test of Direct and Indirect Effects

The hypothesised model was analysed in AMOS, and the results returned a poor

fit, 2 (45) 272.87 p < .001, CMIN/DF = 6.06, CFI = 0.84, TLI = 0.77; NFI = 0.82,

RMSEA = 0.13, SRMR = 0.09. Therefore, the modification indices (MI) were

consulted to examine how the model’s fit could be improved. Covarying the error

terms between the focus and motivation variables would decrease the model

discrepancy by a total of MI = 120.04, while covarying the error terms between educate

and feedback showed a potential reduction of MI = 23.36. Given that these constructs

were conceptually related, covarying the error terms is acceptable. Therefore, these

error terms were covaried and the respecified model fit was improved, 2(41) = 122.69,

p < .001, CMIN/DF = 2.99, NFI = 0.92, TLI = 0.91, CFI = .94, RMSEA = 0.08, SRMR

= .05. The goodness-of-fit indices showed an acceptable fit, and therefore, no other

modifications were conducted. The respecified model is shown in Figure 6.3.


Figure 6.3 Respecified Model after consulting the modification indices

Next, the standardised estimates were examined to see the predictive abilities of

the antecedents and to assess whether the hypotheses were met. As shown in Table

6.3, LMX, anticipation focus, perceived control and felt responsibility were the only

antecedents that showed significant relationships with proactive safety behaviours.

The relationships were positive, indicating that higher levels of LMX, anticipation

focus, perceived control and felt responsibility were associated with higher levels of

proactive safety behaviours. Organisational and group safety climates and LMX did

not demonstrate significant direct effects with proactive safety behaviours.

Table 6.3 Standardised regression weights and 95% Bias-corrected Confidence Intervals with 10,000 Bootstrap sampling

Estimate p-value

BCa 95% CI Lower Interval

BCa 95% CI Upper

Interval SC-Org → Anticipation Focus 0.12 0.120 -0.04 0.28 SC-Group → Anticipation Focus 0.18 0.019 0.03 0.34 SC-Org → Perceived Control 0.16 0.041 0.01 0.31 SC-Group → Perceived Control 0.37 *** 0.21 0.51 SC-Org → Felt Responsibility 0.04 0.621 -0.12 0.20 SC-Group → Felt Responsibility 0.38 *** 0.22 0.52 SC-Org → PSBWD 0.08 0.290 -0.07 0.22 SC-Group → PSBWD -0.07 0.387 -0.23 0.09 LMX → PSBWD 0.14 0.010 0.03 0.25 Anticipation Focus → PSBWD 0.32 *** 0.22 0.42 Perceived Control → PSBWD 0.27 *** 0.13 0.41 Felt Responsibility → PSBWD 0.26 *** 0.13 0.38

Notes. N = 294; *** p = <.001; Bias-corrected.


To test the hypothesised mediation effect of safety climate on proactive safety

behaviours via the focus and motivation variables, the indirect effects were assessed.

The findings showed that a significant indirect effect was only observed for group

safety climate (β = 0.25, BCa 95% CI [.15, .37]; shown as the dotted line in Figure

6.4). This indirect effect was observed via anticipation focus, perceived control and

felt responsibility. No other significant indirect effects were observed.

Figure 6.4 Respecified model with standardised regression estimates, only showing the significant paths.

Hypothesis on the Non-Proactive States, Motivation and Outcomes

Although not the core focus of the current study within the program of research,

it is important to demonstrate that proactive safety behaviours were distinct from safety

compliance behaviour so as to assess the discriminant validity between the two

constructs. Therefore, an additional outcome variable, safety compliance, along with

its proposed antecedent variable, preventative focus, were added to the model to

examine its differential effect on proactive safety behaviours.

When prevention focus and safety compliance were added to the model, the

goodness of fit indices indicated a relatively acceptable fit, 2 (58) 173.56, p < .001,

CMIN/DF = 2.99, NFI = 0.90, TLI = .88, CFI = .93, RMSEA = .08, SRMR = .06.

However, the original model without preventative focus and safety compliance, still

showed the better model fit.


Within this model, prevention focus showed a significant direct effect β = .13,

BCa 95% CI [.02, .24] with safety compliance. It also showed a negative direct effect

with proactive safety behaviours, but the effect was not significant, β = -.03, BCa 95%

CI [-.13, .07]. Significant direct effects were also found between LMX β =. 17, BCa

95% CI [.04, .30] and safety compliance as well as organisational safety climate β =.

29, BCa 95% CI [.14, .43] and safety compliance. When the indirect effects were

examined, no significant effects were found.

6.8 DISCUSSION

The current study (Study 3) examined the hypothesised model of proactive safety

behaviours within the work driving context (PSB-WD) using a large sample of

employees who drive for work. Specifically, it measured the influence of safety

climate, LMX as contextual antecedents, as well as anticipation focus, felt

responsibility and perceived control as proximal antecedents of proactive safety

behaviours. It was hypothesised that safety climate would have a positive direct effect

on proactive work driving safety behaviours and an indirect effect via the proximal

variables. This hypothesis was only partially supported. The findings revealed that

group and organisational safety climate did not show significant direct effects on

proactive behaviours. Only a significant indirect effect was found between group

Figure 6.5 Research model with prevention focus and safety compliance to test the discriminant validity


safety climate and proactive safety behaviours (via anticipation focus, perceived

control and felt responsibility). These results suggest that group safety climate could

influence employees’ perceived control of safety and risk anticipation while driving

for work as well as felt responsibility for work driving safety issues, which in turn,

encourages them to take safety initiatives towards work driving.

On the other hand, organisational safety climate did not show any significant

effect (whether direct or indirect) on work drivers’ proactive safety behaviours, but

showed a significant direct effect on safety climate. The differential effect between

these two levels of safety climate could indicate that, organisational policies and

procedures on work driving safety influence employees’ compliance behaviours. To

go beyond these compliance behaviours, however, employees must perceive that these

organisational policies and procedures on work driving safety are practiced by their

supervisors and group-unit. In other words, even though company policies and

procedures may set the boundaries of the permissible safety practices within the

organisation, the implementation of these practices that take place within the

supervisory teams are more influential among work drivers (Zohar, 2008; Zohar &

Luria, 2005). Organisations may have policies and procedures regarding work driving

safety, but if this is not implemented by supervisors, work drivers may not necessarily

perceive driving safety as important and, therefore, not be proactive about work

driving safety.

It was also hypothesised that LMX would have a positive direct effect on work

drivers’ proactive safety behaviours, which was supported by the findings. This result

is in line with previous research (e.g., Hofmann et al., 2003; Newnam et al., 2012).

Hofmann et al. (2003) found that LMX was significantly related to safety citizenship

behaviours using a military sample. They argued that having high quality relationships

between leaders and members encourage active collaboration on potential safety

issues. Furthermore, Newnam et al. (2012) found a significant relationship between

LMX and safety driving performance on work drivers. Newnam et al. (2012) argued

that a leadership style that promote concern for workers’ well-being are more likely to

engage employees to practice safer work driving behaviours. The current study extends

the knowledge on the role of LMX within occupational safety by demonstrating that,

high quality exchanges between leaders and team members could also encourage

employees to take initiatives towards work driving safety.


Risk anticipation, perceived control and felt responsibility also showed

significant direct effects on proactive safety behaviours. Compared to the contextual

antecedents, these variables also showed the strongest effects, suggesting a more

proximal influence on proactive safety behaviours. The current results provide further

evidence to the role of future orientation and motivational states on proactive safety

behaviours.

More specifically, the findings revealed that anticipation focus has a positive

direct effect on proactive safety behaviours. This result suggests that work drivers who

anticipate risks are more likely to be proactive in managing driving safety. This result

is consistent with previous research by Curcuruto et al. (2015) where they found

anticipation orientation to have positive correlations with safety initiative, voice and

stewardship behaviours. However, their research only examined the bivariate

correlations between these constructs. The current study extends this knowledge by

showing that, using structural modelling technique, anticipation orientation may also

have a possible predictive role on proactive safety behaviours.

Furthermore, the majority of previous studies have conceptualised motivation as

a single construct (see Study 1 in Chapter 4). The current study extends the existing

literature by studying the multi-dimensionality of motivation on work drivers’

proactive safety behaviours. The current findings suggest that employees are

motivated to take initiatives on work driving safety if they perceive themselves as

having influence on the organisation’s safety processes and procedures. In other

words, employees could be proactive regarding work driving safety if they ‘can’ have

a significant impact on current safety practices (i.e., perceived control represents the

‘can-do’ motivational mechanisms; Curcuruto et al., 2016). In addition, employees are

also likely to take initiatives if they feel responsible for their safety (and of their co-

workers’) while driving for work. In other words, work drivers engage in proactive

safety behaviours because they have a ‘reason’ to engage in such behaviours (i.e., felt

responsibility represents the ‘reason-to’ motivational mechanisms; Curcuruto et al.,

2016). The current results extended the work driving literature by demonstrating that,

felt responsibility is not only related to safe driving behaviours (Banks, Davey, &

Biggs, 2010) but also on work drivers’ proactivity towards work driving safety.


6.8.1 Limitations and Future Research

There were a number of limitations that should be noted and directions for future

research. First, while the cross-sectional design employed in the current study is an

important first step toward the understanding of work drivers’ proactive safety

behaviours, longitudinal studies are required to investigate the causal relationships that

were proposed in the model.

Furthermore, the current study did not focus on crashes, near misses and traffic

offences as they are seen as lagging indicators of work driving safety, however, it is

important to acknowledge the non-significant relationships found between these safety

incidents and proactive safety behaviours. These non-significant results could be due

to the small number of participants who reported experiencing crashes, near misses

and traffic offences during working hours. In addition, the crashes and traffic offences

reported by the participant are self-reported which are susceptible to biases (e.g.,

memory recall), in contrast to actual crashes. Future research could further assess the

relationship between proactive safety behaviours and on-road behaviour using other

options such as organisational crash data and naturalistic driving data.

Different organisational levels of analysis (e.g., individual, work teams, top-

management) could also provide a better understanding of how different factors impact

on proactive safety behaviours. For instance, future studies may utilise a multi-level

modelling to further explore the impact of safety climate on different levels of safety

proactivity, within the individual-, team- or organisational-level.

Next, the study utilised self-report measures to assess the proactive safety

behaviours among work drivers. Although self-report measures are often used in traffic

research, these measures are subject to participant biases (Newnam et al., 2012). For

instance, the recruitment procedure of seeking voluntary participation for the survey

may have contributed to selection bias, whereby employees self-selecting into

completing the survey may be already engaging in proactive work driving safety

behaviours. Those who may not be performing such behaviours may be less likely to

volunteer for participation. On a similar note, it is also possible that participants may

respond inaccurately due to social desirability bias, which refers to an individual’s

tendency to portray himself/herself in a favourable light (Nederhof, 1985). Due to

nature of the topic, participants may tailor their answers to demonstrate positive self-

representation. Future studies could attempt to utilise other methods of measuring


proactive work driving safety behaviours to reduce these biases, such as observation

or diary studies.

6.8.2 Chapter Summary and Key Learnings

The current chapter proposed a new model for proactive safety behaviours that

could be applied within the work driving context. This proposed model was then tested

using a sample of work drivers (N = 300) with structural question modelling. The

findings revealed that group safety climate, LMX, anticipation focus, perceived

control and felt responsibility demonstrated significant relationships with work

drivers’ proactive safety behaviours. More specifically, the results revealed that group

safety climate has an indirect effect on work drivers’ proactive safety behaviours via

anticipation focus, perceived control and felt responsibility, while LMX showed a

direct effect on work drivers’ proactive safety behaviours. Anticipation focus,

perceived control and felt responsibility also showed direct effects on work drivers’

proactive safety behaviours. Organisational-level safety climate did not demonstrate

any significant effects on proactive safety behaviours in work driving, but it showed a

significant direct effect on safety compliance behaviours. Preventative focus also

showed a direct effect on safety compliance behaviours, but not on proactive safety

behaviours.

Overall, the current study suggests that when cultivating proactive safety

behaviours in the work driving settings, group safety climate is more important than

organisational safety climate. Relationship between leaders and members is also

crucial in cultivating these behaviours. Furthermore, employees who have high levels

of anticipation risk, perceived control and felt responsibility towards driving safety

may also engage in more proactive safety behaviours while driving for work.

Chapter 7: Discussion and Conclusions 137

Chapter 7: Discussion and Conclusions


The previous chapters presented the program of research that developed and

tested a new measure and a model of proactive safety behaviours within the work

driving context (PSB-WD). This final chapter provides a synthesis of the research

methodology and findings using the research questions as the structure (see Table 7.1).

This chapter also presents the theoretical and practical contributions of the current

program of research, especially within the area of OHS and work driving safety.

Strengths of the research program as well as the limitations and suggestions for future

research are also discussed.

Table 7.1 Overview of design and methodology structured by the research questions

Overarching aim To develop a model and measure of proactive safety behaviours within the work driving context.

RQ1 What do we currently know about the construct of proactive safety behaviour? Study design and methodology:

Literature review Using research from areas of organisational psychology and occupational safety, the concept of proactive safety behaviours and similar constructs were reviewed. Meta-analysis

Review of the current literature on the antecedents of proactive safety behaviours and using meta-analysis to explore the effect of the factors based on the results of 71 studies.

RQ2 What are the proactive safety behaviours that work drivers perform to improve safety (and safety of their co-workers) while driving for work?

Study design and methodology

Item Generation and Expert Panel

Research team for the item generation (n = 3) and road safety experts and supervisors and managers of work drivers for the expert panel (n = 5).

Literature review and nominal group technique process between the research team to identify items for the measure. Online survey for the expert panel to assess content validity and item clarity.

Pilot Testing

Online survey to assess the dimensionality of the PSB-WD scale and to reduce the number of items with 43 work drivers using PCA for the analysis.

Measurement Testing

Online survey to assess the validity of the PSB-WD scale with N = 300 work drivers using EFA and CFA for the analysis.

138 Chapter 7: Discussion and Conclusions

RQ3 What are the individual and organisational factors that promote and inhibit proactive safety behaviours in work drivers?

Study design and methodology

Model Testing

Online survey to assess the hypothesised model with N = 300 work drivers using SEM for the analysis.

7.1.1 Overall Findings in Relation to RQ1

Two reviews were conducted in order to answer the first research question,

“What do we currently know about the construct of proactive safety behaviours?” The

first review critically synthesised and discussed the origins and definitions of proactive

safety behaviours using studies from organisational psychology and general

occupational safety research. The second review involved a meta-analysis conducted

as a research component within the overall program of research. This review explored

the distal and proximal antecedents of proactive safety behaviours that have been

studied within the OHS literature. The overall findings of these two reviews (i.e., the

literature review and the meta-analytical review) are discussed below.

The literature review discussed the existing literature on proactive safety

behaviours. The first account of this construct was found in Andriessen’s (1978) study

on safety behaviours using a sample of construction workers. In addition to complying

with safety rules and regulations, Andriessen (1978) noted that the participants also

engaged in ‘safety initiatives’, which were voluntary behaviours that improved the

safety of their work environment (e.g., voicing safety concerns to their supervisors).

Since then, a plethora of studies have examined these behaviours within the field of

OHS. However, Neal and Griffin’s (2002) research on safety participation, Hoffman

et al.’s (2003) work on safety citizenship and Curcuruto and colleagues’ studies on

safety proactivity (Curcuruto et al., 2015; Curcuruto & Griffin, 2017; Curcuruto,

Mearns, & Mariani, 2016) have provided a more comprehensive understanding of

these behaviours. These evidence-base studies drew upon concepts from the area of

organisational psychology which, in turn, provided theoretical underpinnings for these

behaviours. Specifically, the theoretical underpinnings of proactive safety behaviours

were based on the concepts of contextual performance (Borman & Motowidlo, 1993),

organisational citizenship behaviours (Organ, 1988) and work proactivity (Bateman &

Crant, 1993; Parker et al., 2006).


Several key findings were obtained from the literature review. First, individuals

who engage in more proactive safety behaviours were associated with fewer safety

incidents at work (Hofmann et al., 2003; Neal & Griffin, 2006). These behaviours were

found to be associated with the promotion and improvement of safety within the

workplace, better attitudes towards safety reporting and lower absence rates (e.g.,

Curcuruto & Griffin, 2017; Hofmann et al., 2003). In particular, proactive safety

behaviour was more effective at improving safety outcomes in the longer-term

(measured by reduced accidents and workplace injuries) compared to merely

complying with safety rules and regulations (e.g., Christian et al., 2009; Neal &

Griffin, 2006; Neal et al., 2000).

Second, these behaviours were initially conceptualised as a unidimensional

construct, with most research using Neal and Griffin’s (2000) measure of safety

participation. However, Hofmann et al.’s (2003) research on safety citizenship

expanded the construct of safety participation which provided a broader range of

behaviours. Their research revealed a higher-order construct of safety citizenship

consisting of six dimensions: helping, safety civic virtue (keeping informed), initiating

safety-related change (improving safety), voice, stewardship and whistle-blowing.

More recently, Curcuruto and colleagues (Curcuruto et al. 2015; Curcuruto & Griffin,

2017) built upon Hofmann et al.’s (2003) research. They argued that safety citizenship

could be further distinguished into two categories: prosocial and proactive behaviours.

Curcuruto et al. (2015) argued that proactive safety behaviours should be action-

oriented, change-oriented and future-focused. Prosocial safety behaviours, on the other

hand, are affiliative in nature, and mostly reflect the helping and stewardship

dimensions of organisational citizenship behaviours (Curcuruto et al., 2015).

The synthesis of the literature helped to inform the definition of proactive safety

behaviours, which was then utilised in the current program of research. Specifically,

proactive safety behaviour within the work driving was defined as: 1) behaviours that

improve the context of the work environment to be more supportive of safety -

behaviours that may not directly contribute to workplace safety, but facilitate an

environment that supports safety (i.e., behaviours that create a positive environment

for safety); 2) behaviours that aim to improve workplace safety that cannot be forced

(i.e., self-starting); and 3) behaviours that are change-oriented which aim to improve

the current workplace safety practices.


The review also examined similar concepts of proactive safety behaviours within

the field of general road traffic safety (i.e., outside of the organisational context). From

the review, research on prosocial driving behaviours (Harris et al., 2014) and traffic

safety citizenship (Otto et al., 2016; Finley et al., 2015) tapped into similar constructs

underlying the proposed proactive work driving safety behaviours. Specifically, the

conceptual underpinning of prosocial driving behaviours and traffic safety citizenship

is that these behaviours foster a safer culture within the general road environment. This

conceptual underpinning is similar to an aspect of proactive safety behaviours - to

improve the context of the work environment to be more supportive of safety.

However, prosocial driving behaviour and traffic safety citizenship were missing

critical aspects of proactive safety behaviours (e.g., change-oriented behaviours). For

instance, prosocial driving behaviours often reflect compliance with traffic laws (e.g.,

“obey traffic signs” and “come to a complete stop at a stop sign”; Harris et al., 2014,

p. 3) while traffic safety citizenship behaviours only measure passenger intervening

behaviours (e.g., speaking up when a driver conduct risky behaviours).

In addition to the literature review of proactive safety behaviours and its related

concepts, a meta-analysis was conducted to critically appraise and synthesise the

antecedents of proactive safety behaviours based on 71 studies. Following Curcuruto

and Griffin’s (2017) framework on safety proactivity, this review explored the distal

and proximal factors that could facilitate or inhibit one’s engagement with proactive

safety behaviours. This framework recognised the importance of the work

environment in cultivating these behaviours and the cognitive mechanisms that may

underlie these discretionary behaviours.

Then, the meta-analytical review (i.e., Study 1 of the research program) provided

further evidence that safety climate and leadership styles were important contextual

(i.e., organisational) variables of proactive safety behaviours. Study 1 also further

confirmed that safety knowledge and safety motivation were significant antecedents

of proactive safety behaviours. The effect sizes found for safety knowledge and safety

motivation demonstrated stronger effects, which may indicate a more proximal

relationship with the outcome variable. These findings were consistent with previously

established reviews (Christian et al., 2009; Clarke, 2006, 2013a). However, Study 1

extended the present knowledge by examining the influence of passive leadership,

perceived organisational support, trust, and work design variables on proactive safety


behaviours. More specifically, the findings of the meta-analysis revealed that passive

leadership style could inhibit one’s engagement with proactive safety behaviours as

demonstrated by the significant and negative effect size, while perceived

organisational support, trust and work autonomy were significantly and positively

related to proactive safety behaviours.

Compared with prior reviews which only examined safety knowledge and safety

motivation as proximal antecedents of proactive safety behaviours, the current meta-

analysis investigated a wider range of proximal factors. The findings revealed that

employees’ self-efficacy, perceived control and having a promotion focus showed

significant positive relationships with proactive safety behaviours. When observing

the pattern of the effects, it was evident that the proximal antecedents demonstrated

stronger effects sizes compared to the contextual antecedents. This pattern provided

possible evidence that contextual antecedents are distal in their relationships with

proactive safety behaviours, while the examined cognitive mechanisms could present

a more direct (i.e., proximal) relationship with proactive safety behaviours.

The literature search within the meta-analysis also revealed that, out of the 71

empirical studies that were examined, two studies were found to have some relevance

to the context of proactive work driving safety. Of these two studies, one was

specifically conducted with an aim to improve work driving safety by examining road

safety participation (Keffane, 2015) and another study examined safety voice using a

sample of urban bus drivers (Tucker et al., 2008). These studies, however, only looked

at the unidimensional nature of proactive safety behaviours (road safety participation

and safety voice).

Overall, the findings from the literature review and meta-analysis provided a

comprehensive understanding of proactive safety behaviours and its related concepts.

Both reviews informed the theoretical underpinnings of the model used for the current

program of research. The evidence base also informed the development of the

measurement tool, by adapting established current measures from previous studies to

suit the work driving context. This process resulted in the development of the research

model and the measurement scale of proactive safety behaviours within the work

driving context.



The concept of proactive safety behaviours and its dimensions have not been

extensively studied within the work driving safety context. Thus, the current program

of research developed a new scale to measure work drivers’ proactive safety behaviour

(PSB-WD) and to answer the second research question, “What are the proactive safety

behaviours that work drivers perform to improve safety (and of their co-workers) while

driving for work?”

The process of developing the PSB-WD scale was guided by Hinkin et al.'s

(1997) and DeVellis' (2016) approaches to scale development. Several studies were

conducted to develop the PSB-WD scale. First, an item generation process pooled

existing items from the literature that measured proactive safety behaviours and related

constructs as well as measures of general work proactivity behaviours (Study 2a).

These items were then adapted to suit the work driving safety context. The candidate,

along with her supervisory team, ranked, revised, removed and categorised the pooled

items into possible dimensions of proactive safety behaviours (e.g., feedback inquiry,

helping/volunteerism, changing organisation’s policies and procedures, problem

prevention, stewardship and safety voice). Then, an expert panel comprising road

safety experts, fleet safety managers and supervisors (n = 5) were contacted to provide

further feedback on the newly developed items and the hypothesised categories. Once

the expert panel’s feedback was applied, the scale was piloted with a small sample of

work drivers (n = 43) to assess its dimensionality.

The results of the pilot test (Study 2b) revealed eight possible dimensions of

proactive work driving safety behaviours based on a principal component analysis

(PCA). The results of the PCA also revealed that, generally, most items remained

within the expected factors. For instance, the items for volunteerism/helping, voice

and feedback inquiry were originally written for these factors. However, items written

for the stewardship and problem prevention subscales were mixed. A larger sample of

work drivers (n = 300) was used to further examine the scale by using exploratory and

confirmatory factor analysis (Study 2c). The final dimensions of this construct were

identified as: education, safety voice, fixing safety issues, vehicle maintenance,

intervene, and feedback inquiry. The final PSB-WD scale had 21 items and the

psychometric properties of the PSB-WD scale were demonstrated as having sound

reliability and validity (e.g., discriminant and convergent).


Several key findings were obtained from the scale development process. First,

Study 2 revealed that proactive work driving safety driving behaviour is a multi-

dimensional construct. Next, although the majority of the dimensions resulting from

the scale development have been supported by previous research (e.g., education,

fixing safety issues, feedback inquiry), the newly developed scale provided a specific

dimension that may be unique to work driving context; namely, ‘vehicle maintenance’.

Previous research that examined work drivers’ pre-trip vehicle maintenance

behaviours usually measured this behaviour using two items: “check the tyre pressure

of your work vehicle” and “check the oil and water levels of your work vehicle” (Wills

et al., 2006, p. 377). However, Wills et al. (2006) did not specifically examined the

proactive behaviours of work drivers. Checking the tyre pressure as well as checking

the oil and water levels of work vehicles could only denote safety compliance

behaviours. On the other hand, the vehicle maintenance dimension in the current

program of research provides a specific example of how employees could engage in

behaviours that relate to proactive problem prevention while driving for work. The

items for the vehicle maintenance dimensions also explored the aspect of consulting

another person (e.g., supervisor) when a possible defect occurs.

Last, findings from Study 3 suggested that the current sample of work drivers

reported high levels of engagement in vehicle maintenance behaviours as indicated by

the high average and negative skewed distribution scores within this dimension. On

the other hand, the current sample of work drivers reported low levels of feedback

inquiry as indicated by the low average and positively skewed distribution scores

within this dimension.


The third research question was: “What are the individual and organisational

factors that promote and inhibit proactive safety behaviours in work drivers?” To

answer this research question, a quantitative analysis was conducted to test the

hypothesised model of proactive safety behaviours within the work driving context.

The hypothesised model proposed multi-level safety climate (i.e., organisation-level

and group-level) and leader-member exchange (LMX) as contextual antecedents of

proactive safety behaviours in work drivers; while anticipation focus, perceived

control and felt responsibility were identified as the proximal antecedents. The

findings revealed that organisational and group safety climate were differentially


related to work drivers’ proactive safety behaviours. More specifically, group safety

climate showed a significant indirect effect on proactive safety behaviours, via

anticipation focus, perceived control and felt responsibility. However, organisational

safety climate did not show any significant effect, whether directly or indirectly, on

proactive safety behaviours. These results suggested that group safety climate may be

a more important antecedent of work drivers’ proactive safety behaviours compared

to organisational safety climate.

Furthermore, the relationship between leaders and work drivers (as measured by

LMX) showed a significant positive direct effect on proactive work driving safety

behaviours. Examining the bivariate correlations, LMX demonstrated the strongest

correlation with vehicle maintenance. This finding suggests that high quality

relationships with supervisors could encourage employees to be proactive about their

safety while driving for work, especially when maintaining vehicles. In a practical

manner, these findings suggest that leaders and management could facilitate work

drivers’ safety proactivity by conducting daily or weekly toolbox talks with a specific

focus on work driving safety.

The findings also revealed other factors that significantly related to proactive

safety behaviours and these included anticipation focus, perceived control and felt

responsibility. The current findings suggest that work drivers who anticipate risks, who

perceive that they have control of work driving situations and who feel responsible for

work driving safety, were more likely to be proactive in managing driving safety.

7.1.4 Theoretical Implications and Contributions

The current program of research, while exploratory in nature, presents several

theoretical implications and contributions. First, the development and testing of the

PSB-WD model and scale presents an advancement in knowledge within the OHS and

work driving safety fields. More specifically, even though the construct of proactive

safety behaviours was initially conceptualised within the field of OHS, this concept

has been effectively applied within the field of work driving safety. Furthermore, the

PSB-WD model is the first theoretical framework that considered the organisational

factors and cognitive mechanisms (and the interaction between the two) on proactive

safety behaviours within the work driving context.


On a broad level, the findings of the research program imply that the

management of work driving safety could be advanced by applying the current

knowledge on the fields of OHS and organisational psychology. Since work drivers

operate within the organisational environment (Stuckey et al., 2010), there is an

impetus to examine the role of the work environment in managing the risks related to

work driving. The current program of research also demonstrated that there is a merit

in examining behaviours (other than driving behaviours, and involvement in crashes

and traffic violations) as potential safety performance indicators within the work

driving context.

Another theoretical implication of the current research program is the extension

of safety proactivity within the field of work driving safety. There is an increasing

interest in the role of proactive management of safety within the OHS field (e.g.,

Curcuruto et al., 2015). The current program of research provides a significant

contribution by demonstrating that proactive safety behaviours could be applied within

a specific area of work driving safety.

Put simply, while organisations ensure their employees’ safety by managing

risks and hazards in the workplace, risks related to work driving maybe relatively less

likely to be of focus. For instance, employees in utility industries may undertake

compulsory training to ensure safety while operating maintenance equipment and

machinery. However, utility workers may also drive between different worksites. This

work-related driving activity, which is associated with high risks, is often ignored in

general OHS research (Wishart et al., 2019).

The current research program suggests that the risks associated with these work-

related driving activities should also be managed by organisations, especially given

that driving for work is one of the most hazardous work activities that an employee

could undertake. Future research could also examine the differences in proactive safety

behaviours in both general work activities and when employees must drive for work.

Second, the meta-analysis reported in Study 1 extended the current literature on

proactive safety behaviours. Extending Curcuruto and Griffin's (2017) literature

review and model of safety proactivity, the present study advanced the current

understanding of the contextual and proximal antecedents of safety proactivity by

providing a meta-analytic investigation of various organisational factors and cognitive

mechanisms and their influence on proactive safety behaviours. The meta-analytic


results provide quantitative evidence to support Curcuruto and Griffin’s (2017)

proposed framework for safety proactivity. More specifically, the findings of the meta-

analysis provide further evidence on the importance of examining the multi-level

framework of safety climate, different styles of leadership, social exchanges and

interpersonal processes within the organisation and work design as facilitators of

proactive safety behaviours. The meta-analysis also demonstrated other possible

proximal antecedents of proactive safety behaviours other than safety knowledge and

safety motivation (which has been the focus of previous research). In particular, the

meta-analysis showed that perceived control demonstrated the strongest meta-analytic

effect with proactive safety behaviours, even more than the effects found for safety

knowledge and safety motivation.

Furthermore, the development of the PSB-WD scale provided further evidence

relating to the multi-dimensionality of proactive safety behaviours. The current

program of research also revealed new dimensions which had not been previously

examined. The PSB-WD scale shares similar dimensions to prior measures of

proactive safety behaviours. For instance, voicing driving safety concerns relates to

the voice dimension of Hofmann et al.’s (2002) safety citizenship measure. However,

the PSB-WD scale also demonstrated dimensions of feedback inquiry, fixing safety

issues and vehicle maintenance. These dimensions were specifically developed from

the work proactivity measures by Parker and colleagues (Parker & Collins, 2010;

Parker & Wang; 2015), measures that previous studies on proactive safety behaviours

have not examined. Although these dimensions were specifically developed for work

driving safety, it is possible that these dimensions could also be applied within the

general work safety field.

In addition, the PSB-WD model further confirms the importance of examining

the multi-level framework of safety climate in the work driving setting. The majority

of research within the work driving field does not utilise the multi-level framework of

safety climate (Wills et al., 2006, 2009). It is only recently that Huang and colleagues

(e.g., Huang et al., 2013; Huang et al., 2017) have examined the group- and

organisational-level of safety climate within the work driving context. The current

research program provides further evidence that employees have different perceptions

of safety climate depending on the referent level. For instance, results of Study 3

showed that group safety climate had a significant indirect effect on proactive safety


behaviours (via anticipation orientation, perceived control and felt responsibility).

However, group-level safety climate did not show any significant effects on safety

compliance.

On the other hand, organisational-level safety climate showed a significant direct

effect on safety compliance, but not on proactive safety behaviours. While safety

compliance was not the focus of the current research program, the findings provide

further evidence that different levels of safety climate may have a distinct influence on

different behaviours performed by the work driver (safety compliance versus proactive

safety behaviours). The findings of Study 3 also provided further evidence relating to

the role of leaders on work driving safety, and extended the current literature by

demonstrating that having high quality relationship between leaders and members

could also encourage proactive safety behaviours within the work driving context.

Lastly, the current program of research contributes to the existing literature by

providing more evidence on the utility of examining the cognitive mechanisms that

underlie proactive safety behaviours. Previous research of Curcuruto et al.’s (2015)

only examined the bivariate correlations between future orientation and motivational

states on safety proactivity. Using SEM, the current research program extends this

knowledge by showing that anticipation orientation, perceived control and felt

responsibility demonstrated direct effects on proactive safety behaviours.

7.1.5 Strengths and Practical Implications

Several strengths and practical implications are evident from the research

program, especially relating to the management of work driving safety. Driving for

work is one of the riskiest activity that an employee could undertake during working

hours. It is critical to address the risks associated with work-related driving given that

a high percentage of injuries and fatalities that occur in the work place are vehicle-

related incidents. Although risk management processes relating to high risk industries

are underpinned by the ‘zero harm’ approach, organisations lack policies relating to

driving safety when managing work driving risks. This lack of appropriate risk

management strategies and diligence toward work-related road safety can severely

compromise safe driving practices and subsequently increase the risk, severity and

frequency of work-related crashes (Stuckey et al., 2010). Findings from this program

of research provide significant contributions to the current knowledge within the fields

of OHS and work driving safety by exploring the proactive safety behaviours of


employees who drive for work. Previous research on proactive safety behaviours using

work drivers (e.g., Keffane, 2015 and Tucker et al., 2008) have only examined the

unidimensional nature of proactive safety behaviours (i.e., road safety participation;

Keffane, 2015, and safety voice; Tucker et al., 2008). This is the first research program

that has explored the multi-dimensional concept of proactive safety behaviours within

the work driving field.

The core strength of the current program of research is the comprehensive

understanding of proactive safety behaviours from the literature review and meta-

analysis. Unlike the traditional narrative (unsystematic) reviews in organisational

sciences that only include studies selected by the author, conducting a meta-analytic

review limits this selection bias by thoroughly assembling and critically appraising all

relevant research on the specified topic (Denyer & Tranfield, 2009). The results of the

meta-analysis also further informed the safety proactivity framework that was

developed by Curcuruto and Griffin (2017). Another key strength of the research

program is developing a measurement tool that has psychometrically sound properties.

While there are problems within self-report questionnaires, surveys are still the most

commonly used method in organisational and OHS research as it is cost-efficient and

easy to administer (DeVellis, 2016).

The current program of research also provides significant practical contributions

in the field of work driving safety by 1) investigating how organisations can engage

their work drivers to be more proactive in managing risks while driving for work, and

by 2) providing a potential indicator of work driving safety performance that does not

rely on lagging indicators (e.g., crashes and traffic violations).

The proposed PSB-WD model demonstrated the importance of organisational

factors, particularly group safety climate and LMX, on work drivers’ proactive safety

behaviours. These findings suggest that organisations could encourage these

behaviours among work drivers by ensuring that safety policies and procedures on

work driving are implemented. Direct supervisors’ commitment to safety is especially

important as the current findings revealed that group safety climate may influence

employees’ anticipation of risks when driving for work as well as their perceived

control and felt responsibility in work driving safety. These cognitive mechanisms, in

turn, encourage work drivers to act proactively about safety matters related to work-

related driving. On the other hand, the influence of LMX on work drivers’ proactive


safety behaviours suggests that if supervisors and managers actively demonstrate that

they care and value their employees within the organisation, then employees could

perceive that management as accepting of employees’ work driving safety concerns.

From these results, the following practical suggestions for workplace safety

interventions are discussed. Training could be implemented within organisations to

raise awareness on the influence of safety climate on workers’ safety behaviours while

driving as well as creating leadership programs that promote safety. The current

program of research demonstrated the critical role that supervisors play, suggesting

that leaders need to have high quality relationships with their team members in order

to encourage workers to engage in proactive safety behaviours while driving. These

training programs may include a leadership-based intervention designed to improve

managers’ and supervisors’ interpersonal skills in order to enrich their communication

and relationships with work drivers (Newnam et al., 2012). In addition, organisations

should openly demonstrate that they are committed to the safety of employees when

they drive for work. Organisations must also ensure that work driving safety rules and

procedures do not conflict with requirements of other work tasks to demonstrate

commitment to a strong and positive work driving safety climate.

The results of the Study 3 also showed that employees were motivated to engage

in proactive safety behaviours if they continuously anticipated risks, felt responsible

for their organisation’s work driving issues, and perceived that they have control of

these issues. In a practical manner, these findings suggest that leaders and management

could facilitate work drivers’ safety proactivity by conducting daily or weekly tool-

box talks. Within these meetings, workers and supervisors could be encouraged to

openly communicate possible risks, dangers and hazards that workers may experience

on the road. These meetings could increase employee’s perceived responsibility and

control towards work driving issues, which in turn could motivate them to engage in

proactive safety behaviours. Managers could ask team members one-to-one or in team

meetings to provide feedback or ideas for fleet safety improvement where open

dialogue is encouraged and supported.

Furthermore, the newly developed PSB-WD scale could be used as a diagnostic

tool for work driving safety performance. This psychometrically sound scale could be

used as a complementary tool to current measures of work driving safety (e.g.,

occupational driving safety behaviours, self-reporting of crashes, safety incidents,


traffic violations and near misses). More specifically, instead of just looking at lagging

indicators (e.g., crashes and traffic violations), organisations could use this scale to

measure other indicators of safety performance among work drivers.

The scale could also be used as an education tool to cultivate these behaviours

in organisations. The PSB-WD scale identified various behavioural indicators of

proactive safety within the work driving context. Companies could use the current

research findings to develop workshops which could outline the various ways that

employees could be proactive about safety while driving for work. For instance,

training materials could be developed clearly stating that proactive safety behaviours

are encouraged in the workplace (e.g., “in our company, we ensure workplace safety

by encouraging other drivers to follow safety working procedures”, or “to reduce work

driving incidents, speak up when you see a co-worker using a model phone while

driving”). These statements could also be reiterated by supervisors and could be

incorporated in employee orientations. These trainings should also be conducted on a

regular basis to demonstrate that these behaviours are critical to the organisation.

However, it is important to acknowledge that these training programs may only

be effective if the organisations already operate within a certain level of positive safety

culture regarding work driving (Wishart et al., 2019). In other words, employees must

perceive that these behaviours are valued and encouraged in the workplace. These

behaviours do not occur in a vacuum (Didla et al., 2007; Didla, Mearns, & Flin, 2009),

and the current research showed that safety climate is a critical factor for proactive

safety behaviours within the work driving context. In addition, any training materials

developed based on the results of the current research would need to be evaluated and,

as such, further research would be required to evaluate such interventions.

7.1.6 Limitations and Suggestions for Future Research

The significant theoretical and practical contributions notwithstanding, this

program of research also had limitations which need to be acknowledged. Overall, it

is important to acknowledge that the current program of research was exploratory in

nature given that this was the first study to extensively examine the application of

proactive safety behaviours within the work driving context. Future research directions

are discussed in the following sections.


First, the main survey (Study 2c and Study 3) utilised a convenience sampling

recruitment and, therefore, the results may not be generalised to the whole work

driving population. In addition, the use of self-report measures may be subjected to

self-reporting bias. It is possible that employees who volunteered to complete the

survey were more conscientious about driving safety. Given that the current program

of research is largely exploratory, future research could attempt to utilise other

methods of measuring proactive work driving safety behaviours to reduce these biases,

such as observation method or diary studies. Further, supervisor rates of work team’s

safety proactivity level could offer further evidence of the external validity of the

measure. Future studies could also investigate whether the results found in the current

sample are consistent with employees from other organisations.

Second, due to the cross-sectional design of the main survey that informed the

model testing, data was only collected at a single time-point using one questionnaire.

Single time-point questionnaires have been previously criticised due to their increased

potential for common method variance (e.g., Podsakoff et al. 2003). Although the

current program of research assessed the potential common method bias within dataset,

longitudinal studies are required to minimise the bias that may result from common

method variance. In addition, longitudinal studies could further clarify the

relationships between the hypothesised variables of the PSB-WD model and allow

causality to be inferred. For example, previous research suggests that there is a

reciprocal relationship between safety motivation and safety participation (Neal &

Griffin, 2006). While safety climate is an important antecedent of proactive safety

behaviours, it is also possible that engagement in these behaviours could further

improve the workplace safety climate. Future research could assess the possible

reciprocal relationships. Further re-test reliability of the final PSB-WD scale should

also be conducted.

In addition, while the current program of research did not focus on crashes,

future research could further assess the relationship between proactive safety

behaviours and on-road behaviour using other options such as organisational crash

data and naturalistic driving data. Recently, OHS researchers have begun examining

the influence of proactive safety behaviours on other safety outcome measures such as

lost-time injuries, micro-accidents and property damage (e.g., Curcuruto et al, 2015).

Furthermore, future studies could examine the influence of proactive safety behaviours


on other performance measures that go beyond safety outcomes, such as company

reputation, job satisfaction, increased work engagement (Huang, Lee, McFadden,

Murphy, Robertson, Cheung & Zohar, 2016).

Lastly, the current program of research was primarily conducted using

quantitative methods. For instance, the item pool generation using the current literature

review, nominal group technique and the use of expert panel provided a deductive

approach to the development of the PSB-WD scale. However, using qualitative

methods to obtain opinions gathered from the target population could offer other

insights to the current findings. For instance, future studies could examine other

potential facilitators and barriers to engaging in proactive safety behaviours within the

work driving safety context through an in-depth qualitative investigation.

7.1.7 Concluding Remarks

Existing research and practice on work driving safety mostly focus on lagging

indicators and management of risks is often reactive, only conducting safety checks

once an incident has already occurred. The current program of research made

significant contributions to the current knowledge and industry practice within the

field of OHS by exploring the possible application of the proactive safety behaviour

construct within the context of work driving. This program of research is driven with

the proposition that, to improve safety while driving for work, employees and

management need to be proactive in addressing safety issues as opposed to just simply

following rules and regulations and reacting to crashes, violations and injuries.

Proactive safety behaviour as a measure of safety performance in the work driving

context could provide an alternative or complementary paradigm to current risk

management strategies.

The current exploratory investigation has provided theoretical and practical

contributions to the field of work driving safety by providing a model on how


managing risks while driving for work. Furthermore, the research program also

provided a possible complementary measure that measure behaviour-based leading

indicators of safety for employees who drive for work. Driving for work involves high

risks, thus proactive safety behaviours may be necessary to encourage employees to

engage in activities going beyond compliance. Consequently, engaging in proactive

safety behaviours may vastly improve work driving safety.

Bibliography 153

Bibliography

Abdi, H. (2003). Factor rotations in factor analyses. Encyclopedia for Research

Methods for the Social Sciences. Thousand Oaks, CA: Sage.

Adams-Guppy, J. R., & Guppy, A. (1995). Speeding in relation to perceptions of risk,

utility and driving style by British company car drivers. Ergonomics, 38(12),

2525-2535. doi:10.1080/00140139508925284

Agnew, C., Flin, R., & Mearns, K. J. (2013). Patient safety climate and worker safety

behaviours in acute hospitals in Scotland. Journal of Safety Research, 45, 95-101.

doi:10.1016/j.jsr.2013.01.008

Anderson, J. C., & Gerbing, D. W. (1988). Structural equation modeling in practice:

A review and recommended two-step approach. Psychological Bulletin,

103(3), 411. doi:10.1037/0033-2909.103.3.411

Andriessen, J. (1978). Safe behaviour and safety motivation. Journal of Occupational

Accidents, 1(4), 363-376. doi:10.1016/0376-6349(78)90006-8

Arrigo, G., & Casale, G. (2010). A comparative overview of terms and notions on

employee participation. Retrieved from

https://www.ilo.org/labadmin/info/pubs/WCMS_123713/lang--en/index.htm

Arthur Jr, W., Bell, S. T., Edwards, B. D., Day, E. A., Tubre, T. C., & Tubre, A. H.

(2005). Convergence of self-report and archival crash involvement data: A two-

year longitudinal follow-up. Human factors, 47(2), 303-313. doi:

10.1518/0018720054679416

Aryee, S., & Hsiung, H.-H. (2016). Regulatory focus and safety outcomes: An

examination of the mediating influence of safety behavior. Safety Science, 86,

27-35. doi:10.1016/j.ssci.2016.02.011

Australian Transport Council. (2011). National road safety strategy: 2011-2020.

Retrieved from

https://www.infrastructure.gov.au/roads/safety/national_road_safety_strategy/fi

les/NRSS_2011_2020_15Aug11.pdf.

Austroads. (2018). Vehicles as workplace (1925671313). Retrieved from

https://www.onlinepublications.austroads.com.au/items/AP-R561-18

154 Bibliography

Banks, T. D. (2008). An investigation into how work-related road safety can be

enhanced (Doctoral dissertation). Retrieved from

https://eprints.qut.edu.au/29683/

Banks, T., Davey, J., & Biggs, H. (2010). The influence of safety ownership on

occupational road safety outcomes. Journal of the Australasian College of

Road Safety, 21(4), 36.-42. Retrieved from https://eprints.qut.edu.au/39203/

Barling, J., Loughlin, C., & Kelloway, E. K. (2002). Development and test of a model

linking safety-specific transformational leadership and occupational safety.

Journal of Applied Psychology, 87(3), 488-496. doi: 10.1037/0021-

9010.87.3.488

Bass, B. M. (1997). From transactional to transformational leadership: Learning to

share the vision. Organizational Dynamics, 18(3), 318-333. doi:

10.1016/0090-2616(90)90061-S

Bateman, T. S., & Crant, J. M. (1993). The proactive component of organizational

behavior: A measure and correlates. Journal of Organizational Behavior,

14(2), 103-118. doi:10.1002/job.4030140202

Beus, J. M., Dhanani, L. Y., & McCord, M. A. (2015). A meta-analysis of personality

and workplace safety: Addressing unanswered questions. Journal of Applied

Psychology, 100(2), 481-498. doi: 10.1037/a0037916

Borman, W. C., & Motowidlo, S. (1993). Expanding the criterion domain to include

elements of contextual performance. In N. Schmitt & W.C. Borman (Eds.),

Personnel Selection in Organizations (pp. 71-98). San Francisco: Jossey-Bass.

Brondino, M., Silva, S. A., & Pasini, M. (2012). Multilevel approach to organizational

and group safety climate and safety performance: Co-workers as the missing

link. Safety Science, 50(9), 1847-1856. doi:10.1016/j.ssci.2012.04.010

Brown, T. A. (2014). Confirmatory factor analysis for applied research. New York,

NY: Guilford Publications.

Buckley, L., Sheehan, M., & Chapman, R. (2009). Adolescent protective behavior to

reduce drug and alcohol use, alcohol-related harm and interpersonal violence.

Journal of drug education, 39(3), 289-301. doi:10.2190/de.39.3.e

Bureau of Labor Statistics. (2014). Census of Fatal Occupational Injuries Charts,

1992-2014 (preliminary data). Retrieved from

http://www.bls.gov/iif/oshcfoi1.htm.

Bibliography 155

Burke, M. J., Sarpy, S. A., Tesluk, P. E., & Smith-Crowe, K. (2002). General safety

performance: A test of a grounded theoretical model. Personnel Psychology,

55(2), 429-457. doi: 10.1111/j.1744-6570.2002.tb00116.x

Burt, C. D., Sepie, B., & McFadden, G. (2008). The development of a considerate and

responsible safety attitude in work teams. Safety Science, 46(1), 79-91. doi:

10.1016/j.ssci.2006.10.005

Byrne, B. M. (2016). Structural equation modeling with AMOS: Basic concepts,

applications, and programming. New York, NY: Routledge.

Chen, C. F., & Chen, S. C. (2014). Investigating the effects of job demands and job

resources on cabin crew safety behaviors. Tourism Management, 41, 45-52.

doi:10.1016/j.tourman.2013.08.009

Christian, M. S., Bradley, J. C., Wallace, J. C., & Burke, M. J. (2009). Workplace

safety: a meta-analysis of the roles of person and situation factors. Journal of

Applied Psychology, 94(5), 1103-1127. doi:10.1037/a0016172

Chughtai, A. A. (2015). Creating safer workplaces: The role of ethical leadership. Safety

Science, 73, 92-98. doi:10.1016/j.ssci.2014.11.016

Clarke, S. (2006). The relationship between safety climate and safety performance: A

meta-analytic review. Journal of Occupational Health Psychology, 11(4), 315-

327. doi:10.1037/1076-8998.11.4.315

Clarke, S. (2013a). Safety leadership: A meta-analytic review of transformational and

transactional leadership styles as antecedents of safety behaviours. Journal of

Occupational and Organizational Psychology, 86(1), 22-49.

doi:10.1111/j.2044-8325.2012.02064.x

Clarke, S. (2013b). Safety leadership: A meta-analytic review of transformational and

transactional leadership styles as antecedents of safety behaviours. Journal of

Occupational and Organizational Psychology, 86(1), 22-49. doi:

10.1111/j.2044-8325.2012.02064.x

Clarke, S., & Ward, K. (2006). The Role of Leader Influence Tactics and Safety Climate in

Engaging Employees' Safety Participation. Risk Analysis, 26(5), 1175-1185.

doi:10.1111/j.1539-6924.2006.00824.x

Conchie, S. M. (2013). Transformational leadership, intrinsic motivation, and trust: A

moderated-mediated model of workplace safety. Journal of Occupational

Health Psychology, 18(2), 198-210. doi:10.1037/a0031805

Conchie, S. M. (2013). Transformational leadership, intrinsic motivation, and trust: A

156 Bibliography

moderated-mediated model of workplace safety. Journal of Occupational Health

Psychology, 18(2), 198-210. doi:10.1037/a0031805

Conchie, S. M., & Donald, I. J. (2008). The functions and development of safety-

specific trust and distrust. Safety Science, 46(1), 92-103.

doi:http://dx.doi.org/10.1016/j.ssci.2007.03.004

Conchie, S. M., & Donald, I. J. (2009). The moderating role of safety-specific trust on

the relation between safety-specific leadership and safety citizenship

behaviors. Journal of Occupational Health Psychology, 14(2), 137-147.

doi:10.1037/a0014247

Conchie, S. M., & Donald, I. J. (2009). The moderating role of safety-specific trust on the

relation between safety-specific leadership and safety citizenship behaviors. Journal

of Occupational Health Psychology, 14(2), 137-147. doi:10.1037/a0014247

Conchie, S. M., Donald, I. J., & Taylor, P. J. (2006). Trust: Missing piece (s) in the

safety puzzle. Risk Analysis, 26(5), 1097-1104. doi: 10.1111/j.1539-

6924.2006.00818.x

Conchie, S. M., Taylor, P. J., & Donald, I. J. (2012). Promoting safety voice with safety-

specific transformational leadership: The mediating role of two dimensions of trust.

Journal of Occupational Health Psychology, 17(1), 105-115. doi: 10.1037/a0025101

Copsey, S., Rommel, A., Cleal, Y., Kärmeniemi, P., Stengård, J., Blobner, K., . . .

Kouvonen, A. (2011). Managing risks to drivers in road transport. Retrieved

from https://osha.europa.eu/en/tools-and-

publications/publications/reports/managing-risks-drivers_TEWE11002ENN

Craciun, G., Shin, D., & Zhang, J. Q. (2017). Safe driving communication: A

regulatory focus perspective. Journal of Consumer Behaviour, 16(6), e50-e60.

doi: 10.1002/cb.1654

Cropanzano, R., & Mitchell, M. S. (2005). Social exchange theory: An

interdisciplinary review. Journal of management, 31(6), 874-900. doi:

10.1177/0149206305279602

Curcuruto, M., & Griffin, M. A. (2017). Safety proactivity in the workplace: The

initiative to improve individual, team, and organizational safety. In S. K. Parker

& U. K. Bindl (Eds.), Proactivity at Work: Making things happen in

organizations. New York, NY: Routledge.

Curcuruto, M., Conchie, S. M., Mariani, M. G., & Violante, F. S. (2015). The role of

prosocial and proactive safety behaviors in predicting safety performance. Safety

Bibliography 157

Science, 80, 317-323. doi:10.1016/j.ssci.2015.07.032

Curcuruto, M., Mearns, K. J., & Mariani, M. G. (2016). Proactive role-orientation

toward workplace safety: Psychological dimensions, nomological network and

external validity. Safety Science, 87, 144-155. doi:10.1016/j.ssci.2016.03.007

Dabos, G. E., & Rousseau, D. M. (2004). Mutuality and reciprocity in the

psychological contracts of employees and employers. Journal of applied

psychology, 89(1), 52-72. doi:10.1037/0021-9010.89.1.52

DeArmond, S., Smith, A. E., Wilson, C. L., Chen, P. Y., & Cigularov, K. P. (2011).

Individual safety performance in the construction industry: Development and

validation of two short scales. Accident Analysis and Prevention, 43(3), 948-954.

doi:10.1016/j.aap.2010.11.020

Delbecq, A. L., Van de Ven, A. H., & Gustafson, D. H. (1975). Group techniques for

program planning: A guide to nominal group and Delphi processes. Glenview,

IL: Scott, Foresman and Company.

Department of Infrastructure Regional Development and Cities. (2018). Road Safety.

Retrieved from https://infrastructure.gov.au/roads/safety/

DeVellis, R. F. (2016). Scale development: Theory and applications (Vol. 26): New

York, NY: Sage Publications.

Didla, S., Mearns, K., & Flin, R. (2007). Safety citizenship behaviour in the oil and

gas industry. In T. Aven, J.E. Vinnem (Eds.), Risk, Reliability and Societal

Safety. London: Taylor & Francis Group.

Didla, S., Mearns, K., & Flin, R. (2009). Safety citizenship behaviour: A proactive

approach to risk management. Journal of Risk Research, 12(3-4), 475-483.

doi:10.1080/13669870903041433

Efron, B., & Tibshirani, R. (1986). Bootstrap methods for standard errors, confidence

intervals, and other measures of statistical accuracy. Statistical Science, 1(1)

54-75. Retrieved from https://projecteuclid.org/euclid.ss/1177013815

Eisenberger, R., Huntington, R., Hutchison, S., & Sowa, D. (1986). Perceived

organizational support. Journal of Applied Psychology, 71(3), 500-507.

European Agency for Safety and Health at Work. (2010). A review of accidents and

injuries to road transport drivers. Retrieved from

https://osha.europa.eu/en/tools-and-

publications/publications/literature_reviews/Road-transport-accidents.pdf.

158 Bibliography

European Agency for Safety and Health at Work. (2012). Worker participation

practices: A review of EU-OSHA case studies. Retrieved from

https://osha.europa.eu/en/publications/literature_reviews/worker-

participation-practices-a-review-of-eu-osha-case-studies/view

Field, A. (2013). Discovering statistics using IBM SPSS statistics. New York, NY:

Sage.

Finley, K., Riggs, S., Otto, J., & Ward, N. J. (2015). An assessment of traffic safety

culture: exploring traffic safety citizenship. Retrieved from

https://trid.trb.org/view/1378597

Ford, M. T., & Tetrick, L. E. (2011). Relations among occupational hazards, attitudes, and

safety performance. Journal of Occupational Health Psychology, 16(1), 48-66.

doi:10.1037/a0021296

Fort, E., Ndagire, S., Gadegbeku, B., Hours, M., & Charbotel, B. (2016). Working

conditions and occupational risk exposure in employees driving for work.

Accident Analysis & Prevention, 89, 118-127. doi:10.1016/j.aap.2016.01.015

Fugas, C. S., Meliá, J. L., & Silva, S. A. (2011). The 'is' and the 'ought': How do perceived

social norms influence safety behaviors at work? Journal of Occupational Health

Psychology, 16(1), 67-79. doi:10.1037/a0021731

Fugas, C. S., Melia, J. L., & Silva, S. A. (2011). The "Is" and the "Ought": How Do

Perceived Social Norms Influence Safety Behaviors at Work? Journal of

Occupational Health Psychology, 16(1), 67-79. doi:10.1037/a0021731

Fugas, C. S., Silva, S. A., & Melia, J. L. (2012). Another look at safety climate and

safety behavior: Deepening the cognitive and social mediator mechanisms.

Accident Analysis and Prevention, 45, 468-477. doi:10.1016/j.aap.2011.08.013

Fugas, C. S., Silva, S. A., & Melia, J. L. (2012). Another look at safety climate and safety

behavior: Deepening the cognitive and social mediator mechanisms. Accident

Analysis and Prevention, 45, 468-477. doi:10.1016/j.aap.2011.08.013

Fugas, C. S., Silva, S. A., & Melia, J. L. (2013). Profiling Safety Behaviors:

Exploration of the Sociocognitive Variables that Best Discriminate Between

Different Behavioral Patterns. Risk Analysis, 33(5), 838-850.

doi:10.1111/j.1539-6924.2012.01913.x

Gaskin. (2011). Excel StatTools. Retrieved from

http://statwiki.kolobkreations.com/index.php?title=Main_Page

Bibliography 159

Geller, E. S., Johnson, C., Redmon, W., & Mawhinney, T. (2001). Actively caring for

occupational safety: Extending the performance management paradigm. In

C.M. Johnson, W.K. Redmon, T.C. Mawhinney (Eds.), Handbook of

organizational performance: Behavior analysis and management (pp. 303-

326). New York, NY: Haworth.

Geller, E. S., Roberts, D. S., & Gilmore, M. R. (1996). Predicting propensity to

actively care for occupational safety. Journal of Safety Research, 27(1), 1-8.

doi:10.1016/0022-4375(95)00024-0

Geyskens, I., Krishnan, R., Steenkamp, J.-B. E., & Cunha, P. V. (2009). A review and

evaluation of meta-analysis practices in management research. Journal of

Management, 35(2), 393-419. doi: 10.1177/0149206308328501

Glendon, I. (2011). Safety and Risk in Transportation. In R. J. Burke, S. Clarke, & C.

L. Cooper (Eds.), Occupational health and safety (pp. 239-275). London, UK:

Gower Publishing, Ltd.

Grant, A. M., & Ashford, S. J. (2008). The dynamics of proactivity at work. Research

in organizational behavior, 28, 3-34. doi: 10.1016/j.riob.2008.04.002

Grant, A. M., & Parker, S. K. (2009). 7 redesigning work design theories: the rise of

relational and proactive perspectives. The Academy of Management Annals,

3(1), 317-375. doi: 10.5465/19416520903047327

Griffin, M. A., & Curcuruto, M. (2016). Safety climate in organizations: New

challenges and frontiers for theory, research and practice. Annual Review of

Organizational Psychology and Organizational Behavior, 3(1), 11:11-11:22.

doi: 10.1146/annurev-orgpsych-041015-062414

Griffin, M. A., & Neal, A. (2000). Perceptions of safety at work: A framework for

linking safety climate to safety performance, knowledge, and motivation.

Journal of Occupational Health Psychology, 5(3), 347-358. doi:10.1037/1076-

8998.5.3.347

Griffin, M., & Talati, Z. (2014). Safety leadership The Oxford handbook of leadership and

organizations (pp. 638-656): Oxford University Press, USA.

Guldenmund, F. W. (2000). The nature of safety culture: a review of theory and

research. Safety science, 34(1), 215-257. doi:10.1016/s0925-7535(00)00014-x

Haddad, H., & Delhomme, P. (2006). Persuading young car drivers to take part in a

driving skills test: The influence of regulatory fit on informational-assessment

160 Bibliography

value and persuasion. Transportation Research Part F: Traffic Psychology and

Behaviour, 9(6), 399-411. doi: 10.1016/j.trf.2006.02.002

Hair, J. F., William, C. B., Babin, B. J., & Anderson, R. E. (2010). Multivariate data

analysis. London, UK: Pearson Education LTD.

Hale, A., & Borys, D. (2013). Working to rule or working safely? Part 2: The

management of safety rules and procedures. Safety Science, 55, 222-231.

doi:10.1016/j.ssci.2012.05.013

Hales, D. N., & Chakravorty, S. S. (2016). Creating high reliability organizations using

mindfulness. Journal of Business Research, 69(8), 2873-2881. doi:

10.1016/j.jbusres.2015.12.056

Harris, P. B., Houston, J. M., Vazquez, J. A., Smither, J. A., Harms, A., Dahlke, J. A.,

& Sachau, D. A. (2014). The prosocial and aggressive driving inventory

(PADI): a self-report measure of safe and unsafe driving behaviors. Accident


Haworth, N., Greig, K., & Wishart, D. (2008). Improving fleet safety: Current

approaches and best practice guidelines (AP-R321/08). Retrieved from


Haworth, N., Tingvall, C., & Kowadlo, N. (2000). Review of best practice road safety

initiatives in the corporate and/or business environment. Retrieved from

http://www.monash.edu/__data/assets/pdf_file/0007/216484/muarc166.pdf

Health and Safety Executive. (2014). Driving at work: Managing work-related road

safety. Retrieved from http://www.hse.gov.uk/pubns/indg382.pdf

Hinkin, T. R. (1998). A brief tutorial on the development of measures for use in survey

questionnaires. Organizational Research Methods, 1(1), 104-121.

doi:10.1177/109442819800100106

Hinkin, T. R., Tracey, J. B., & Enz, C. A. (1997). Scale construction: Developing

reliable and valid measurement instruments. Journal of Hospitality & Tourism

Research, 21(1), 100-120. doi:10.1177/109634809702100108

Hoffmeister, K., Gibbons, A. M., Johnson, S. K., Cigularov, K. P., Chen, P. Y., &

Rosecrance, J. C. (2014). The differential effects of transformational leadership

facets on employee safety. Safety Science, 62, 68-78. doi:

10.1016/j.ssci.2013.07.004

Hofmann, D. A., & Morgeson, F. P. (1999). Safety-related behavior as a social

exchange: The role of perceived organizational support and leader–member

Bibliography 161

exchange. Journal of Applied Psychology, 84(2), 286-296. doi: 10.1037/0021-

9010.84.2.286

Hofmann, D. A., & Morgeson, F. P. (2004). The role of leadership in safety. In J. E.

Barling & M. R. Frone (Eds.), The psychology of workplace safety (pp. 159-

180). Washington, DC: American Psychological Association.

Hofmann, D. A., Morgeson, F. P., & Gerras, S. J. (2003). Climate as a moderator of

the relationship between leader-member exchange and content specific

citizenship: Safety climate as an exemplar. Journal of Applied Psychology,

88(1), 170-178. doi:10.1037/0021-9010.88.1.170

Hofmann, D. A., Morgeson, F. P., & Gerras, S. J. (2003). Climate as a moderator of the

relationship between leader-member exchange and content specific citizenship:

Safety climate as an exemplar. Journal of Applied Psychology, 88(1), 170-178.

doi:10.1037/0021-9010.88.1.170

Hogan, J., & Foster, J. (2013). Multifaceted personality predictors of workplace safety

performance: More than conscientiousness. Human performance, 26(1), 20-

43. doi:10.1080/08959285.2012.736899

Hon, C. K. H., Chan, A. P. C., & Yam, M. C. H. (2014). Relationships between safety

climate and safety performance of building repair, maintenance, minor alteration, and

addition (RMAA) works. Safety Science, 65, 10-19. doi:10.1016/j.ssci.2013.12.012

Howard, M. C. (2016). A review of exploratory factor analysis decisions and overview

of current practices: What we are doing and how can we improve?

International Journal of Human-Computer Interaction, 32(1), 51-62. doi:

10.1080/10447318.2015.1087664

Hu, L., & Bentler, P. M. (1995). Evaluating Model Fit. In R. H. Hoyle (Ed.), Structural

equation modeling: Concepts, issues, and applications (pp. 76-88). New York,

NY: Sage.

Hu, X., Yeo, G., & Griffin, M. (2018). Safety compliance re-examined: Differentiating

deep compliance from surface compliance. Paper presented at the Academy of

Management Proceedings. Retrieved from

https://journals.aom.org/doi/abs/10.5465/AMBPP.2018.15338abstract

Huang, Y. H., Lee, J., McFadden, A. C., Murphy, L. A., Robertson, M. M., Cheung,

J. H., & Zohar, D. (2016). Beyond safety outcomes: An investigation of the

impact of safety climate on job satisfaction, employee engagement and

162 Bibliography

turnover using social exchange theory as the theoretical framework. Applied

ergonomics, 55, 248-257. doi: 10.1016/j.apergo.2015.10.007

Huang, Y.-H., Ho, M., Smith, G. S., & Chen, P. Y. (2006). Safety climate and self-

reported injury: Assessing the mediating role of employee safety control.

Accident Analysis & Prevention, 38(3), 425-433. doi:

10.1016/j.aap.2005.07.002

Huang, Y.-h., Lee, J., McFadden, A. C., Rineer, J., & Robertson, M. M. (2017).

Individual employee’s perceptions of “Group-level Safety Climate”

(supervisor referenced) versus “Organization-level Safety Climate” (top

management referenced): Associations with safety outcomes for lone workers.

Accident Analysis & Prevention, 98, 37-45. doi: 10.1016/j.aap.2016.09.016.

Huang, Y.-H., Zohar, D., Robertson, M. M., Garabet, A., Murphy, L. A., & Lee, J.

(2013). Development and validation of safety climate scales for mobile remote

workers using utility/electrical workers as exemplar. Accident Analysis &

Prevention, 59, 76-86. doi:10.1016/j.aap.2013.04.030

Hudson, P. (2003). Applying the lessons of high risk industries to health care. Quality

and Safety in Health care, 12 Suppl 1(suppl 1), i7-12.

doi:10.1136/qhc.12.suppl_1.i7

Inness Michelle, M., Turner, N., Barling, J., & Stride, C. B. (2010). Transformational

leadership and employee safety performance: A within-person, between-jobs design.

Journal of Occupational Health Psychology, 15(3), 279-290. doi:10.1037/a0019380

Jiang, L., & Probst, T. M. (2016). Transformational and passive leadership as cross-

level moderators of the relationships between safety knowledge, safety

motivation, and safety participation. Journal of Safety Research, 57, 27-32.

doi:10.1016/j.jsr.2016.03.002

Jiang, L., Yu, G., Li, Y., & Li, F. (2010). Perceived colleagues' safety knowledge/behavior

and safety performance: Safety climate as a moderator in a multilevel study. Accident

Analysis and Prevention, 42(5), 1468-1476. doi:10.1016/j.aap.2009.08.017

Jolliffe, I. T., & Cadima, J. (2016). Principal component analysis: a review and recent

developments. Philosophical Transactions A, 374(2065), 20150202. doi:

10.1098/rsta.2015.0202

Kark, R., Katz-Navon, T., & Delegach, M. (2015). The dual effects of leading for safety:

The mediating role of employee regulatory focus. Journal of Applied Psychology,

100(5), 1332. doi: 10.1037/a0038818

Bibliography 163

Kath, L. M., Marks, K. M., & Ranney, J. (2010). Safety climate dimensions, leader–member

exchange, and organizational support as predictors of upward safety communication

in a sample of rail industry workers. Safety Science, 48(5), 643-650. doi:

10.1016/j.ssci.2010.01.016

Kedjidjian, C. (1995). Fleet safety: How to save lives and money. Retrieved from


Keffane, S. (2015). Communication’s role in safety management and performance of

the road safety practices. Jordan Journal of Civil Engineering, 9(2), 229-244.

doi: 10.1260/2046-0430.3.1.79

Kelloway, E. K., Mullen, J., & Francis, L. (2006). Divergent effects of

transformational and passive leadership on employee safety. Journal of

Occupational Health Psychology, 11(1), 76. doi: 10.1037/1076-8998.11.1.76

Kennedy, A., Cullen, B., Firman, D., Fleiter, J. J., & Lewis, I. (2018). Peer passenger

intentions to speak up to a risky driver: A theoretically-guided investigation of

the effects of a high school road safety education program. Transportation

Research Part F: Traffic Psychology and Behaviour, 54, 15-27. doi:

10.1016/j.trf.2018.01.011

Kinnear, N., Kelly, S. W., Stradling, S., & Thomson, J. (2013). Understanding how drivers

learn to anticipate risk on the road: A laboratory experiment of affective anticipation

of road hazards. Accident Analysis & Prevention, 50, 1025-1033. doi:

10.1016/j.aap.2012.08.008

Kline, P. (2000). Handbook of psychological testing (2nd ed.). New York, NY:

Routledge.

Krause, N., Ragland, D. R., Greiner, B. A., Syme, L., & Fisher, J. M. (1997).

Psychosocial job factors associated with back and neck pain in public transit

operators. Scandinavian Journal of Work, Environment & Health, 23(3), 179-

186. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/9243727

Kurtessis, J. N., Eisenberger, R., Ford, M. T., Buffardi, L. C., Stewart, K. A., & Adis,

C. S. (2017). Perceived organizational support: A meta-analytic evaluation of

organizational support theory. Journal of management, 43(6), 1854-1884. doi:

10.1177/0149206315575554

Lekka, C. (2011). High reliability organisations: A review of the literature. Retrieved

from http://www.hse.gov.uk/research/rrpdf/rr899.pdf

Lewis-Beck, M., Bryman, A. E., & Liao, T. F. (2003). The Sage encyclopedia of social

164 Bibliography

science research methods. New York, NY: Sage Publications.

Li, F., Wang, G., Li, Y., & Zhou, R. (2017). Job demands and driving anger: The roles

of emotional exhaustion and work engagement. Accident Analysis &

Prevention, 98, 198-205. doi:https://doi.org/10.1016/j.aap.2016.10.013

Lindell, M. K., & Whitney, D. J. (2001). Accounting for common method variance in

cross-sectional research designs. Journal of Applied Psychology, 86(1), 114.

doi: 10.1037//0021-9010.86.1.114

Little, T. D., Cunningham, W. A., Shahar, G., & Widaman, K. F. (2002). To parcel or

not to parcel: Exploring the question, weighing the merits. Structural equation

modeling, 9(2), 151-173. doi: 10.1207/S15328007SEM0902_1

Lofquist, E. A. (2010). The art of measuring nothing: The paradox of measuring safety

in a changing civil aviation industry using traditional safety metrics. Safety

Science, 48(10), 1520-1529. doi:10.1016/j.ssci.2010.05.006

Lofquist, E. A. (2017). Jousting with dragons: A Resilience Engineering approach to

managing SMS in the transport sector. Retrieved from https://www.itf-

oecd.org/sites/default/files/docs/resilience-engineering-approach-managing-

sms.pdf

Matsunaga, M. (2008). Item parceling in structural equation modeling: A primer.

Communication Methods and Measures, 2(4), 260-293. doi:

10.1080/19312450802458935

Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995). An integrative model of

organizational trust. Academy of Management Review, 20(3), 709-734.

Retrieved from

https://www.jstor.org/stable/258792?seq=1#metadata_info_tab_contents

Mearns, K. J., & Flin, R. (1999). Assessing the state of organizational safety—culture

or climate? Current Psychology, 18(1), 5-17. doi: 10.1007/s12144-999-1013-

3

Mearns, K. J., & Reader, T. (2008). Organizational support and safety outcomes: An

un-investigated relationship? Safety Science, 46(3), 388-397.

doi:10.1016/j.ssci.2007.05.002

Mooren, L. (2016). An evidence-based safety management system for heavy truck

transport operations (Doctoral dissertation). Retrieved from


Bibliography 165

Morrow, P. C., & Crum, M. R. (2004). Antecedents of fatigue, close calls, and crashes

among commercial motor-vehicle drivers. Journal of Safety Research, 35(1),

59-69. doi: 10.1016/j.jsr.2003.07.004

Murray, W. (2007). Worldwide Occupational Road Safety (WORS) Review Project.

Retrieved from http://eprints.qut.edu.au/7143/1/7143a.pdf

Murray, W., Newnam, S., Watson, B. C., Davey, J., & Schonfeld, C. C. (2003).

Evaluating and improving fleet safety in Australia. Retrieved from

https://infrastructure.gov.au/roads/safety/publications/2003/pdf/eval_fleetsafe

.pdf.

Murray, W., Watson, B., King, M., Pratt, S., & Darby, P. (2014). Applying the Haddon

matrix in the context of work-related road safety. Retrieved from

https://eprints.qut.edu.au/78733/1/OSIT14_Session2A_Watson.pdf

Nahrgang, J. D., Morgeson, F. P., & Hofmann, D. A. (2011). Safety at work: A meta-

analytic investigation of the link between job demands, job resources, burnout,

engagement, and safety outcomes. Journal of Applied Psychology, 96(1), 71-

94. doi: 10.1037/a0021484

National Safety Council. (2014). Risk perception: Theories, strategies, and next steps.

Retrieved from

https://www.nsc.org/Portals/0/Documents/CambpellInstituteandAwardDocu

ments/WP-Risk%20Perception.pdf

Naveh, E., & Marcus, A. (2002). A re-assessment of road accident data-analysis

policy: applying theory from involuntary, high-consequence, low-probability

events like nuclear power plant meltdowns to voluntary, low-consequence,

high-probability events like traffic accidents. Retrieved from

http://www.cts.umn.edu/Publications/ResearchReports/reportdetail.html?id=5

20

Neal, A., & Griffin, M. A. (2002). Safety climate and safety behaviour. Australian

journal of management, 27(1 suppl), 67-75.

doi:10.1177/031289620202701S08

Neal, A., & Griffin, M. A. (2006). A study of the lagged relationships among safety

climate, safety motivation, safety behavior, and accidents at the individual and

group levels. Journal of Applied Psychology, 91(4), 946. doi:10.1037/0021-

9010.91.4.946

166 Bibliography

Neal, A., Griffin, M. A., & Hart, P. M. (2000). The impact of organizational climate

on safety climate and individual behavior. Safety Science, 34(1), 99-109. doi:

10.1016/S0925-7535(00)00008-4

Neubert, M. J., Kacmar, K. M., Carlson, D. S., Chonko, L. B., & Roberts, J. A. (2008).

Regulatory focus as a mediator of the influence of initiating structure and

servant leadership on employee behavior. Journal of Applied Psychology,

93(6), 1220. doi: 10.1037/a0012695

Newnam, S., & Goode, N. (2015). Do not blame the driver: A systems analysis of the causes

of road freight crashes. Accident Analysis & Prevention, 76, 141-151. doi:

10.1016/j.aap.2015.01.016

Newnam, S., & Oxley, J. (2016). A program in safety management for the occupational

driver: Conceptual development and implementation case study. Safety Science, 84,

238-244. doi: 10.1016/j.ssci.2015.12.020

Newnam, S., & VonSchuckmann, C. (2012). Identifying an appropriate driving

behaviour scale for the occupational driving context: The DBQ vs. The ODBQ.

Safety Science, 50(5), 1268-1274. doi:10.1016/j.ssci.2011.12.009

Newnam, S., & Watson, B. (2011a). A comparison of the driving behavior between

remunerated and volunteer drivers. Safety Science, 49(2), 339-344.

doi:10.1016/j.ssci.2010.09.012

Newnam, S., & Watson, B. (2011b). Work-related driving safety in light vehicle fleets:

A review of past research and the development of an intervention framework.


Newnam, S., Goode, N., Salmon, P., & Stevenson, M. (2017). Reforming the road freight

transportation system using systems thinking: An investigation of Coronial inquests

in Australia. Accident Analysis & Prevention, 101, 28-36.

Newnam, S., Greenslade, J., Newton, C., & Watson, B. (2011). Safety in occupational

driving: Development of a driver behavior scale for the workplace context.

Applied Psychology: An International Review, 60(4), 576-599.

doi:10.1111/j.1464-0597.2011.00448.x

Newnam, S., Griffin, M. A., & Mason, C. (2008). Safety in work vehicles: A multilevel

study linking safety values and individual predictors to work-related driving

crashes. Journal of Applied Psychology, 93(3), 632-644. doi: 10.1037/0021-

9010.93.3.632

Bibliography 167

Newnam, S., Lewis, I., & Watson, B. (2012). Occupational driver safety:

conceptualising a leadership-based intervention to improve safe driving

performance. Accident Analysis & Prevention, 45, 29-38.

doi:10.1016/j.aap.2011.11.003

Newnam, S., Sheppard, D. M., Griffin, M. A., McClure, R. J., Heller, G., Sim, M. R.,

& Stevenson, M. R. (2014). Work-related road traffic injury: A multilevel

systems protocol. Injury prevention, 20(4), e6-e6. doi: 10.1136/injuryprev-

2013-041132

Newnam, S., Warmerdam, A., Sheppard, D., Griffin, M., & Stevenson, M. (2017). Do

management practices support or constrain safe driving behaviour? A multi-

level investigation in a sample of occupational drivers. Accident Analysis &

Prevention, 102, 101-109. doi: 10.1016/j.aap.2017.02.007

Newnam, S., Watson, B., & Murray, W. (2004). Factors predicting intentions to speed

in a work and personal vehicle. Transportation Research Part F: Traffic

Psychology and Behaviour, 7(4), 287-300. doi: 10.1016/j.trf.2004.09.005

Newnam, S., Wishart, D., & Newton, J. (2014). Fleet safety benchmarking: stage 1

literature summary, and stage 2 consultation and framework development:

research report for the national road safety partnership program. Retrieved

from http://www.nrspp.org.au/Pool/Resources/NRSPP-Stage-2-National-

Benchmark-Project-Final-Report-2015-06(1).pdf

O’Neill, S., Martinov-Bennie, N., Cheung, A., & Wolfe, K. (2013). Issues in the

measurement and reporting of work health and safety performance: A review.

Retrieved from

https://www.safeworkaustralia.gov.au/system/files/documents/1703/issues-

measurement-reporting-whs-performance.pdf

Organ, D. W. (1988). Organizational citizenship behavior: The good soldier

syndrome. Lexington, MA: Lexington Books.

Organ, D. W., Podsakoff, P. M., & MacKenzie, S. B. (2006). Organizational

citizenship behavior: its nature, antecedents, and consequences. Thousand

Oaks: SAGE Publications.

Otto, J., Finley, K., & Ward, N. J. (2016). An assessment of traffic safety culture

related to engagement in efforts to improve traffic safety. Retrieved from

https://westerntransportationinstitute.org/wp-

168 Bibliography

content/uploads/2018/01/4W5482_SAFETY_CITIZENSHIP_FINAL_REPO

RT.pdf

Otto, J., Ward, N., Swinford, S., & Linkenbach, J. (2014). Engaging worksite

bystanders to reduce risky driving. Transportation Research Part F: Traffic

Psychology and Behaviour, 26, 370-378. doi: 10.1016/j.trf.2014.02.006

Parker, S. K., & Collins, C. G. (2010). Taking stock: Integrating and differentiating

multiple proactive behaviors. Journal of management, 36(3), 633-662.

Parker, S. K., & Wang, Y. (2015). Helping people to ‘make things happen’: A

framework for proactivity at work. International Coaching Psychology

Review, 10(1), 63-75. Retrieved from

https://www.researchgate.net/publication/272821927_Helping_people_to_'m

ake_things_happen'_A_framework_for_proactivity_at_work

Parker, S. K., Axtell, C. M., & Turner, N. (2001). Designing a safer workplace:

Importance of job autonomy, communication quality, and supportive

supervisors. Journal of Occupational Health Psychology, 6(3), 211-228.

Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/11482633

Parker, S. K., Williams, H. M., & Turner, N. (2006). Modeling the antecedents of

proactive behavior at work. Journal of Applied Psychology, 91(3), 636.

doi:10.1037/0021-9010.91.3.636

Peretz, R. A., & Luria, G. (2017). Drivers’ social-work relationships as antecedents of

unsafe driving: A social network perspective. Accident Analysis & Prevention,

106, 348-357. doi: 10.1016/j.aap.2017.07.005

Podsakoff, P. M., MacKenzie, S. B., Lee, J.-Y., & Podsakoff, N. P. (2003). Common

method biases in behavioral research: A critical review of the literature and

recommended remedies. Journal of Applied Psychology, 88(5), 879. doi:

10.1037/0021-9010.88.5.879

Poulter, D. R., Chapman, P., Bibby, P. A., Clarke, D. D., & Crundall, D. (2008). An

application of the theory of planned behaviour to truck driving behaviour and

compliance with regulations. Accident Analysis & Prevention, 40(6), 2058-

2064. doi: 10.1016/j.aap.2008.09.002

Pratt, S. G. (2003). Work-related roadway crashes: Challenges and opportunities for

presentation. Cincinnati, OH: Centers for Disease Control and Prevention, The

National Institute for Occupational Safety and Health.

Bibliography 169

Queensland Government. (2011). Work Health and Safety Act 2011. Retrieved from

https://www.worksafe.qld.gov.au/laws-and-compliance/workplace-health-

and-safety-laws/laws-and-legislation/work-health-and-safety-act-2011

Raggatt, P. T. (1991). Work stress among long-distance coach drivers: A survey and

correlational study. Journal of organizational behavior, 12(7), 565-579. doi:

10.1002/job.4030120702

Rhoades, L., & Eisenberger, R. (2002). Perceived organizational support: a review of

the literature. Journal of Applied Psychology, 87(4), 698-714. doi:

10.1037/0021-9010.87.4.698

Robb, G., Sultana, S., Ameratunga, S., & Jackson, R. (2008). A systematic review of

epidemiological studies investigating risk factors for work-related road traffic

crashes and injuries. Injury prevention, 14(1), 51-58.

doi:10.1136/ip.2007.016766

Roberts, D. S., & Geller, E. S. (1995). An "actively caring" model for occupational

safety: A field test. Applied and Preventive Psychology, 4(1), 53-59.

doi:10.1016/S0962-1849(05)80051-4

Roberts, D. S., & Geller, E. S. (2001). An “actively caring” model for occupational

safety. In E. S. Geller & J. H. Williams (Eds.), Keys to Behavio-Based Safety

(Vol. 4, pp. 71-83). Rockville, MD: ABS Consulting, Government Institutes.

Safe Work Australia. (2013). Compendium of workers' compensation statistics:

Australia 2010-11. Retrieved from

https://www.safeworkaustralia.gov.au/system/files/documents/1702/compend

ium_2010-11.pdf

Safe Work Australia. (2016a). Guide to the model work health and safety act.

Retrieved from

https://www.safeworkaustralia.gov.au/system/files/documents/1702/guide-to-

the-whs-act-at-21-march-2016.pdf

Safe Work Australia. (2016b). Worker presentation and participation guide.

Safe Work Australia. (2018a). Fatality statistics. Retrieved from

https://www.safeworkaustralia.gov.au/statistics-and-

research/statistics/fatalities/fatality-statistics

Safe Work Australia. (2018b). Work-related injury fatalities - Key WHS statistics

Australia 2018. Retrieved from

170 Bibliography

https://www.safeworkaustralia.gov.au/book/work-related-injury-fatalities-

key-whs-statistics-australia-2018

Salminen, S. (2000). Traffic accidents during work and work commuting.

International Journal of Industrial Ergonomics, 26(1), 75-85. doi:

10.1016/S0169-8141(00)00003-2

Salmon, P. M., & Lenné, M. G. (2015). Miles away or just around the corner? Systems

thinking in road safety research and practice. Accident; analysis and prevention, 74,

243. doi: 10.1016/j.aap.2014.08.001

Salmon, P. M., McClure, R., & Stanton, N. A. (2012). Road transport in drift? Applying

contemporary systems thinking to road safety. Safety Science, 50(9), 1829-1838. doi:

10.1016/j.ssci.2012.04.011

Sampson, J. M., DeArmond, S., & Chen, P. Y. (2014). Role of safety stressors and

social support on safety performance. Safety Science, 64, 137-145.

doi:10.1016/j.ssci.2013.11.025

Schafer, J. L., & Graham, J. W. (2002). Missing data: our view of the state of the art.

Psychological methods, 7(2), 147-177. Retrieved from

https://www.ncbi.nlm.nih.gov/pubmed/12090408

Schmidt, F. L., & Hunter, J. E. (2014). Methods of meta-analysis: Correcting error

and bias in research findings: New York, NY: Sage publications.

Simard, M., & Marchand, A. (1994). The behaviour of first-line supervisors in accident

prevention and effectiveness in occupational safety. Safety Science, 17(3), 169-

185. doi: 10.1016/0925-7535(94)90010-8

Sleet, D. A., Dinh-Zarr, T. B., & Dellinger, A. M. (2007). Traffic safety in the context

of public health and medicine. Retrieved from https://trid.trb.org/view/809702

Stuckey, R., Lamontagne, A. D., Glass, D. C., & Sim, M. R. (2010). Occupational light

vehicle use: Characterising the at-risk population. Journal of health and safety

research and practice, 2(1), online-online. doi:10.1179/oeh.2005.11.2.167

Stuckey, R., Pratt, S. G., & Murray, W. (2013). Work-related road safety in Australia,

the United Kingdom and the United States of America: an overview of

regulatory approaches and recommendations to enhance strategy and

practice. Journal of the Australasian College of Road Safety, 24(3), 10-20

Tabachnick, B. G., & Fidell, L. S. (2007). Using multivariate statistics: Boston, MA:

Pearson Education.

Bibliography 171

Tucker, S., Chmiel, N., Turner, N., Hershcovis, M. S., & Stride, C. B. (2008).

Perceived organizational support for safety and employee safety voice: the

mediating role of coworker support for safety. Journal of Occupational Health

Psychology, 13(4), 319. doi: 10.1037/1076-8998.13.4.319.

Turner, N., Chmiel, N., & Walls, M. (2005). Railing for safety: job demands, job

control, and safety citizenship role definition. Journal of Occupational Health

Psychology, 10(4), 504. doi: 10.1037/1076-8998.10.4.504

Turner, N., Stride, C. B., Carter, A. J., McCaughey, D., & Carroll, A. E. (2012). Job

Demands-Control-Support model and employee safety performance. Accident


Useche, S. A., Ortiz, V. G., & Cendales, B. E. (2017). Stress-related psychosocial

factors at work, fatigue, and risky driving behavior in bus rapid transport (BRT)

drivers. Accident Analysis & Prevention, 104, 106-114. doi:

10.1016/j.aap.2017.04.023

Viechtbauer, W. (2010). Metafor: Meta-analysis package for R. R package version,

2010, 1-0.

Wallace, C., & Chen, G. (2006). A multilevel integration of personality, climate, self-

regulation, and performance. Personnel Psychology, 59(3), 529-557. doi:

10.1111/j.1744-6570.2006.00046.x

Warmerdam, A., Newnam, S., Sheppard, D., Griffin, M., & Stevenson, M. (2017). A

new approach to managing work-related road traffic injury: The development

of a health investment framework. Traffic Injury Prevention, 18(6), 631-635.

doi: 10.1080/15389588.2017.1288289

Warmerdam, A., Newnam, S., Sheppard, D., Griffin, M., & Stevenson, M. (2017).

Workplace road safety risk management: An investigation into Australian practices.

Accident Analysis & Prevention, 98, 64-73. doi: 10.1016/j.aap.2016.09.014

Warmerdam, A., Newnam, S., Wang, Y., Sheppard, D., Griffin, M., & Stevenson, M.

(2018). High performance workplace systems’ influence on safety attitudes

and occupational driver behaviour. Safety Science, 106, 146-153. doi:

10.1016/j.ssci.2018.03.016

Werth, L., & Förster, J. (2007). The effects of regulatory focus on braking speed.

Journal of Applied Social Psychology, 37(12), 2764-2787. doi:

10.1111/j.1559-1816.2007.00280.x

172 Bibliography

Willis, P. G., Brown, K. A., & Prussia, G. E. (2012). Does employee safety influence

customer satisfaction? Evidence from the electric utility industry. Journal of

Safety Research, 43(5-6), 389-396. doi:10.1016/j.jsr.2012.10.003

Wills, A., Watson, B., & Biggs, H. (2006). Comparing safety climate factors as

predictors of work-related driving behavior. Journal of Safety Research, 37(4),

375-383. doi:10.1016/j.jsr.2006.05.008

Wills, A., Watson, B., & Biggs, H. (2009). An exploratory investigation into safety

climate and work-related driving. Work: Journal of Prevention, Assessment &

Rehabilitation, 32(1), 81-94. doi: 10.3233/WOR-2009-0818

Wishart, D. (2015). The challenge of developing a fleet driving risk assessment tool:

What can be learned from the process? (Doctoral dissertation). Retrieved from


Wishart, D., & Davey, J. D. (2004). A research based case study approach to the

development of fleet safety interventions in large vehicle fleets. Paper presented

at the Safety in Action Conference, Melbourne, Australia.

Wishart, D., Bevan, R., & Somoray, K. (in press2019). Safety Citizenship Behaviour:

A complimentary paradigm to improving safety culture within the

organisational driving setting. In N. J. Ward, B. Watson, & T. Özkan (Eds.),

Traffic Safety Culture: Theory, Measurement and Application.

WorkCover Queensland. (2016). Journey claims. Retrieved from

https://www.worksafe.qld.gov.au/rehab-and-claims/injuries-at-work/what-

happens-after-a-claim-is-made/journey-claims

World Health Organization. (2015). World report on road traffic injury prevention.

Retrieved from

https://www.who.int/violence_injury_prevention/publications/road_traffic/wo

rld_report/en/

Xerri, M. (2013). Workplace relationships and the innovative behaviour of nursing

employees: a social exchange perspective. Asia Pacific Journal of Human

Resources, 51(1), 103-123. doi:10.1111/j.1744-7941.2012.00031.x

Zacharatos, A., Barling, J., & Iverson, R. D. (2005). High-Performance Work Systems

and Occupational Safety. Journal of Applied Psychology, 90(1), 77-93.

doi:10.1037/0021-9010.90.1.77

Bibliography 173

Zhang, J., & Wu, C. (2014). The influence of dispositional mindfulness on safety

behaviors: A dual process perspective. Accident Analysis and Prevention, 70,

24-32. doi:10.1016/j.aap.2014.03.006

Zohar, D. (1980). Safety climate in industrial organizations: theoretical and applied

implications. Journal of Applied Psychology, 65(1), 96. doi:10.1037//0021-

9010.65.1.96

Zohar, D. (2002). The effects of leadership dimensions, safety climate, and assigned

priorities on minor injuries in work groups. Journal of organizational

behavior, 23(1), 75-92.

Zohar, D. (2008). Safety climate and beyond: A multi-level multi-climate framework.


Zohar, D. (2010). Thirty years of safety climate research: Reflections and future

directions. Accident Analysis & Prevention, 42(5), 1517-1522.

doi:10.1016/j.aap.2009.12.019

Zohar, D. (2014). Safety climate: Conceptualization, measurement, and improvement.

In B. Schneider & K.M Barbera, The Oxford handbook of organizational

climate and culture (pp. 317-334). Oxford, UK: Oxford Handbooks.

Zohar, D., & Luria, G. (2005). A multilevel model of safety climate: cross-level

relationships between organization and group-level climates. Journal of

Applied Psychology, 90(4), 616-628. doi: 10.1037/0021-9010.90.4.616

Zohar, D., Huang, Y.-h., Lee, J., & Robertson, M. (2014). A mediation model linking

dispatcher leadership and work ownership with safety climate as predictors of

truck driver safety performance. Accident Analysis & Prevention, 62, 17-25.

doi: 10.1016/j.aap.2013.09.005

Zohar, D., Huang, Y.-h., Lee, J., & Robertson, M. M. (2015). Testing extrinsic and

intrinsic motivation as explanatory variables for the safety climate–safety

performance relationship among long-haul truck drivers. Transportation

Research Part F: Traffic Psychology and Behaviour, 30, 84-96. doi:

https://doi.org/10.1016/j.trf.2015.01.014

Appendices 175

Appendices

Appendix A: Coding of Contextual and Proximal Antecedents

Antecedent Variable coded

Safety climate Safety climate, safety culture, perceived safety norms

Psychological safety climate No mentioned level

Group-level Group, co-worker, supervisor, employee, leaders

Organisational-level Organisational

Leadership

Transformational Transformational, LMX, motivate, concern, empowering, safety-specific transformational, ethical

Active transactional

Transactional, monitoring, instructing, safety-specific transformational

Passive Passive leadership

Perceived organisational support

Perceived organisational support, group process, interpersonal, teamwork, organisational concern for employees, supervisor support, positive and non-related job communication, social support, co-worker support

Organisational trust Trust, trust in management

Work Demands Risk, work overload, role overload, work hazards, hazard exposure, emotional demands, job demands

Work Autonomy Job autonomy, job control (but used autonomy as a definition)

Safety Motivation Safety motivation, expectation motivation, motivation, participation motivation, motivation to participate, intrinsic motivation

Safety Knowledge Safety knowledge, knowledge-related job characteristics, sufficiency of technical skills, knowledge

Regulatory focus

Promotion-focused Promotion focus, identified regulation, consideration of future consequences

Prevention-focused Prevention focus, external regulation

Self-efficacy Self-efficacy, perceived probability of success, core self-evaluation

Perceived control Perceived manageability of risks, empowerment, perceived behavioural control, safety control

176 Appendices

Appendix B: Proactive safety behaviours of work drivers item development

Definition of proactive safety behaviours:

1) Behaviours that create a work environment that promotes safety 2) Behaviours that aim to improve workplace safety that is self-starting (in

other words, cannot be forced). 3) Behaviours that are change-oriented which aims to improve the current

workplace safety practices Definition of work drivers: employees who drive for work at least once a week, including commute to and from work. Proactive safety behaviours within the work driving safety context: Behaviours that work drivers perform that create a work environment that promotes work driving safety. These behaviours must be self-starting (in other words, cannot be forced) and/or change-oriented (with an aim to improve the current work driving safety practices). The research team (Klaire Somoray, Darren Wishart and Cameron Newton) have identified SIX subscales that reflect this definition of proactive safety behaviours. These dimensions are: feedback inquiry, helping/volunteerism, changing the organisations’ policies and procedures, problem prevention, stewardship and voice (See figure below). The items that we have generated or revised from previous studies should reflect these dimensions.

Proactive Safety

Behaviours

Feedback Inquiry


Changing Organisation's

Policies and Procedures

Problem Prevention

Stewardship

Voice

Appendices 177

INSTRUCTIONS FOR THE EXPERT PANEL: The following items were developed from previous research on proactive safety behaviours and similar constructs. The items have been revised to suit the work-driving context. We would like to ask you to evaluate these items depending on their content validity and the clarity of meaning. Please rate the content validity of each item, on the scale of 1 (low content validity) to 5 (high content validity) under the heading CV. Please also rate the clarity of each item on the scale of 1 (not clear) to 5 (very clear) under the heading Clarity. If you would like to add some comments regarding each item, please write them within the comments column. If you would like to add general comments, a wider comment section is provided at the end of each subscale (e.g., if there’s an additional item you can think of that fit this subscale, how the items should be re-worded, etc)

SUBSCALE: FEEDBACK INQUIRY Feedback Inquiry involves explicit verbal requests for feedback regarding the employees’ safety behaviours while driving for work. CV Clarity Comments 1 I actively seek feedback from my supervisor

about my work driving

2 I actively seek feedback from my co-workers about my work driving

3 I actively seek feedback about my work driving 4 When I experience a traffic incident or receive a

traffic offence when driving for work, I initiate the conversation to learn from my mistakes

5 I ask people at work about my driving General comments:

SUBSCALE: VOLUNTEERISM/HELPING Volunteering to carry out activities that are not formally part of their job as well as helping and cooperating with others to ensure the safety of other employees while driving for work. CV Clarity Comments 1 I volunteer to educate work driving safety

procedures to new drivers

2 I assist my co-workers to help them drive safely 3 I get involved in work driving safety activities to

help other workers drive safely

4 I volunteer to help other drivers learn more about safe work driving practices

178 Appendices

5 I volunteer to help other drivers with their work safety driving responsibilities

General comments:

SUBSCALE: CHANGING ORGANISATION’S POLICIES AND PROCEDURES Taking one’s initiative to change the organisational policies and procedures with an aim to improve the safety of work drivers. CV Clarity Comments 1 I take the initiative in improving work driving

safety procedures and policies

2 I take the initiative in setting up work driving safety objectives and/or improvement plans

3 I take the initiative in bringing about improved work driving safety procedures within my organisation

4 I try to change the way the job is done to make driving for work safer

5 I try to change work driving safety policies and procedures to make them safer

6 I provide suggestions to my supervisors to improve the safety of the employees while driving for work

7 I make suggestions to improve how work driving safety is handled in our organisation

8 I try to introduce new work driving safety policies that are more effective

9 I provide suggestions to improve employees' safety while driving for work

General comments:

SUBSCALE: PROBLEM PREVENTION Acting to prevent the re-occurrence of challenges and barriers to safety while driving for work CV Clarity Comments 1 When driving for work, I plan extra journey time

and breaks for bad weather, traffic congestion, etc.

2 When driving for work, I plan ahead to make sure I do not feel stressed when driving for work

3 I try to solve problems in ways that reduce risks associated with driving for work

Appendices 179

4 I solve potential issues that relate to driving for work

5 I try to implement solutions to solve urgent work driving safety issues

6 When I see a potential work driving safety hazard, I correct it myself if possible

7 If I see something unsafe, I go out of my way to take care of it

8 I fix safety issues that relates to work driving even if it is not my responsibility

9 If I notice a defect in the vehicle I am driving, I take an appropriate action by notifying my supervisor or complete paperwork

10 When I see a vehicle that needs maintenance, I inform my supervisors about it

11 I volunteer to conduct work driving risk assessments

12 I volunteer to conduct journey planning 13 When I experience a near-miss event, I talk to

people at work to learn from the incident

General comments:

SUBSCALE: STEWARDSHIP Helping co-workers to ensure their safety while driving for work CV Clarity Comments 1 If I see my co-workers doing something risky

while driving for work (e.g., using their mobile phone or speeding), I talk to them about the hazards of their risky behaviours

2 I confront other coworkers about their unsafe behaviours while driving for work

3 I take action to stop safety violations of other drivers to improve work driving safety

4 I warn other coworkers about the dangers of driving unsafely

5 I inform new drivers that violations of safety procedures will not be tolerated

6 I talk to other drivers to follow safe work driving procedures

7 I encourage new drivers to follow safe working procedures

8 I go out of my way to look out for the safety of other drivers

9 I take action to protect other drivers from risky situations

10 I try to prevent other drivers from being injured on the job

General comments:

180 Appendices

SUBSCALE: VOICE Speaking up when you have work driving safety concerns CV Clarity Comments 1 I speak up with new ideas or changes in work

driving safety

2 I speak up and get involved with work driving safety issues that affect work drivers

3 I speak up about safety concerns during team meetings or toolbox talks

4 I speak up on work driving safety matters even if others disagree

5 I communicate my views about work driving safety issue, even if others would disagree

General comments:

Appendices 181

Appendix C: Participant Information Sheet for Pilot Study

PARTICIPANT INFORMATION FOR QUT RESEARCH PROJECT – Pilot Survey –

Beyond Compliance: The investigation of Proactive Safety Behaviours in the Work Driving

Context

QUT Ethics Approval Number 1600000824

RESEARCH TEAM Principal Researcher: Ms Klaire Somoray PhD student Associate Researchers: Dr Darren Wishart Principal Supervisor

Centre for Accident Research and Road Safety – Queensland (CARRS-Q)

Professor Cameron Newton Associate Supervisor School of Management, QUT Business School

Queensland University of Technology (QUT) DESCRIPTION This project is being undertaken as part of a PhD study investigating the proactive safety behaviours performed by workers in organisations that operate vehicle fleets. The purpose of this project is to examine the proactive safety behaviours within the work-related driving setting. Proactive safety behaviours are behaviours that individuals perform to ensure and/or to improve their safety and their coworkers’ while driving for work. These behaviours are above and beyond the employees’ official job’s description and organisational policies and procedures. The project aims: To understand what are the proactive safety behaviours and initiatives that workers engage in to improve their safety and of their co-workers’ while driving for work, that are beyond their formal role. To understand the factors that motivate and hinder proactive safety behaviours at work. To understand the consequences of engaging in proactive safety behaviours at work. You are invited to participate in this project because: Driving is an essential part of your job duties and You drive for work at least once a week. Supervisors, managers and work health and safety officers who also deals with work drivers as part of their jobs are also invited to participate in this project. PARTICIPATION Participation will involve completing an anonymous survey that will take approximately 20 to 30 minutes of your time.

182 Appendices

Questions will include: I get involved in work driving safety activities to help other workers drive safely I take the initiative in improving work driving safety procedures and policies I make suggestions to improve how work driving safety is handled in our organisation Your participation in this project is entirely voluntary. If you agree to participate, you do not have to complete any question(s) you are uncomfortable answering. Your decision to participate or not participate will not impact upon your current or future relationship with QUT, CARRS-Q or your organisation. If you do agree to participate, you can withdraw from the project without comment or penalty. However, as the survey is anonymous once it has been submitted it will not be possible to withdraw. EXPECTED BENEFITS It is expected that this project may not directly benefit you, but it may benefit the general community. The overall aim of the project is to improve employees’ safety while driving for work by understanding the proactive safety behaviours that work drivers may engage in and how we can encourage these behaviours in the workplace. RISKS There are minimal risks associated beyond normal day-to-day living associated with your participation in this project. The questions on this survey will focus on positive and proactive behaviours in the workplace as well as questions about your organisations’ safety climate, safety leadership, motivation and risk perception. If there are items that you are not comfortable in answering, you have the option to skip them. All information you provide will be treated as confidential and will not be shared. Any reports, findings or papers produced using data collected during this research project will only contain aggregate data (i.e. participants grouped together) in order to ensure no individual or identifying details are included. QUT provides for limited free psychology, family therapy or counselling services (face-to-face only) for research participants of QUT projects who may experience discomfort or distress as a result of their participation in the research. Should you wish to access this service please call the Clinic Receptionist on 07 3138 0999 (Monday–Friday only 9am–5pm), QUT Psychology and Counselling Clinic, 44 Musk Avenue, Kelvin Grove, and indicate that you are a research participant. Alternatively, Lifeline provides access to online, phone or face-to-face support, call 13 11 14 for 24 hour telephone crisis support. PRIVACY AND CONFIDENTIALITY All comments and responses are anonymous and will be treated confidentially unless required by law. The names of individual persons are not required in any of the responses. Any data collected as part of this project will be stored securely as per QUT’s Management of research data policy. Please note that non-identifiable data from this project may be used as comparative data in future projects or stored on an open access database for secondary analysis. CONSENT TO PARTICIPATE The return of the completed survey via mail or online is accepted as an indication of your consent to participate in this project.

Appendices 183

QUESTIONS / FURTHER INFORMATION ABOUT THE PROJECT If you have any questions or require further information please contact one of the listed researchers. Klaire Somoray [email protected] 07 3138 4909 Darren Wishart [email protected] 07 3138 4885 Cameron Newton [email protected] 07 3138 2523 CONCERNS / COMPLAINTS REGARDING THE CONDUCT OF THE PROJECT QUT is committed to research integrity and the ethical conduct of research projects. However, if you do have any concerns or complaints about the ethical conduct of the project you may contact the QUT Research Ethics Advisory Team on 07 3138 5123 or email [email protected]. The QUT Research Ethics Advisory Team is not connected with the research project and can facilitate a resolution to your concern in an impartial manner. THANK YOU FOR HELPING WITH THIS RESEARCH PROJECT. PLEASE PRINT THIS SHEET FOR YOUR INFORMATION.

184 Appendices

Appendix D: Proactive Safety Behaviours Pilot Survey

PROACTIVE SAFETY BEHAVIOURS AND WORK-RELATED DRIVING SAFETY

In partnership with your organisation, the Centre for Accident Research and Road Safety Queensland (CARRS-Q) at Queensland University of Technology (QUT) is currently doing a research project on work-related driving safety. The major component of the project is to understand the positive and proactive behaviours that work drivers and their supervisors perform to prevent accidents and unsafe practice while driving for work. The survey is strictly confidential and the research team will combine all the collected responses from the survey, therefore, your specific response will be anonymous. The survey should only take 20 to 30 minutes to complete. THANK YOU FOR YOUR ASSISTANCE. If you would like to participate in the study, consent is required. Please click the button below whether you would like to provide your consent to participate in this study. I would like to provide my conset to participate in this study (1) I do not want to participate in the study (2) INSTRUCTIONS For the following questions, please select the answer which BEST reflect your views and/or experiences. THE SURVEY WILL BE OPEN FOR 2 WEEKS, please complete the survey within this timeframe. Thank you.

Appendices 185

PLEASE NOTE: "Driving for work" includes driving for the purposes of work-related tasks (e.g., going to meetings, driving to see a client or a customer, driving to a work site for a job). This also INCLUDES commuting between work and home. Approximately, how many hours PER WEEK do you normally drive for work?

o 1 – 10 hours (1)

o 11 – 20 hours (2)

o 21 – 30 hours (3)

o 31 – 40 hours (4)

o 41 – 50 hours (5)

o 51 – 60 hours (6)

o 61 hours or more (7)

o I do not drive for work (8) Approximately, how many kilometres do you drive EACH YEAR for work?

o 1 – 10,000 kms (1)

o 10,001 - 20,000 kms (2)

o 20,001 - 30,000 kms (3)

o 30,001 - 40,000 kms (4)

o 40,001 - 50,000 kms (5)

o 50,001 - 60,000 kms (6)

o 60,001 kms or more (7)

o I do not drive for work (8)

186 Appendices

The following questions will ask some information about you and your work vehicle use. Are you?

o Male (1)

o Female (2) What is your age?

o Under 18 years old (1)

o 19 – 24 years old (2)

o 25 – 29 years old (3)

o 30 – 34 years old (4)

o 35 – 39 years old (5)

o 40 – 44 years old (6)

o 45 – 49 years old (7)

o 50 and above (8) How long have you had your driver’s licence?

o Less than 1 year (1)

o 1 – 5 years (2)

o 6 – 10 years (3)

o 11 – 15 years (4)

o 15 – 20 years (5)

o More than 20 years (6)

Appendices 187

How long have you been working for this organisation?

o Less than 1 year (1)

o 1 to 3 years (2)

o 4 to 6 years (3)

o More than 7 years (4) Which department of your organisation do you work for? What type of vehicle do you mainly drive for work?

o Car / Sedan / Wagon (1)

o SUV / 4WD (2)

o Heavy Vehicle (3)

o Other (please specify) (4) ________________________________________________

When driving for work, what type of road do you usually drive on?

o Asphalt / Bitumen roads (1)

o Dirt roads (2)

o Combination of Asphalt / Bitumen roads and Dirt roads (3)

o Other (please specify) (4) ________________________________________________

188 Appendices

The following questions will ask about your exposure to work-related driving risks. Before answering the questions, please read the definitions below: ‘Driving for work’ - includes driving for the purposes of work-related tasks. This INCLUDES commuting between work and home. ‘Crashes’ - any incident involving a motor vehicle that resulted in an injury or any damage to a vehicle or property. 'Near-miss' - any incident involving a motor vehicle where you narrowly avoided a collision with another vehicle or property. ‘Offences’ - any incident for which you were fined or incurred a loss of demerit points. This DOES NOT INCLUDE parking offences. In the last 12 months, how many crashes have you experienced while driving for work?

o None (1)

o One crash (2)

o Two crashes (3)

o Three or more crashes (4) In the last 12 months, how many near-misses have you experienced while driving for work?

o None (1)

o One near-miss (2)

o Two near-misses (3)

o Three or more near-misses (4) In the last 12 months, how many traffic offences have you experienced while driving for work?

o None (1)

o One traffic offence (2)

o Two traffic offences (3)

o Three or more traffic offences (4)

Appendices 189

The following questions will ask you about some behaviours that you may perform at work.

Nev

er

So

met

imes

Ab

ou

t h

alf

of

the

tim

e

Mo

st o

f th

e

tim

e

Alw

ays

I actively seek feedback from my supervisor about my work driving. 1 2 3 4 5

I actively seek feedback from my co-workers and passengers about my work driving. 1 2 3 4 5

I actively seek feedback about my work driving with people at work. 1 2 3 4 5

When I experience a traffic incident or receive a traffic offence when driving for work, I initiate the conversation to learn from my mistakes. 1 2 3 4 5

I ask people at work about my driving. 1 2 3 4 5

I volunteer to educate work driving safety procedures to new drivers. 1 2 3 4 5

I assist my co-workers to help them drive safely. 1 2 3 4 5

I get involved in work driving safety programs and activities to help other workers drive safely. 1 2 3 4 5

I volunteer to help other workers learn more about safe work driving practices. 1 2 3 4 5

I volunteer to help other drivers with their work safety driving responsibilities. 1 2 3 4 5

I take the initiative in improving work driving safety policies and procedures. 1 2 3 4 5

I take the initiative in setting up work driving safety objectives and/or improvement plans within my organisation. 1 2 3 4 5

190 Appendices

I take the initiative in bringing about improved work driving safety procedures within my organisation. 1 2 3 4 5

I provide suggestions to my supervisors to improve the safety of the employees while driving for work. 1 2 3 4 5

I change work driving safety policies and procedures to make them safer. 1 2 3 4 5

I make suggestions to improve how work driving safety is handled in our organisation. 1 2 3 4 5

I introduce new work driving safety policies that are more effective. 1 2 3 4 5

I provide suggestions to improve employees' safety while driving for work. 1 2 3 4 5

When driving for work, I plan extra journey time and breaks for bad weather, traffic congestion, etc. 1 2 3 4 5

I plan ahead to make sure I do not feel stressed when driving for work. 1 2 3 4 5

I resolve problems in ways that reduce the risks associated with driving for work. 1 2 3 4 5

I implement solutions to solve safety issues that relate to driving for work. 1 2 3 4 5

When I see a potential work driving safety hazard, I do my best to fix it. 1 2 3 4 5

If I see something unsafe, I go out of my way to take care of it. 1 2 3 4 5

I fix safety issues that relates to work driving even if it is not my responsibility. 1 2 3 4 5

If I notice a defect in the vehicle I am driving, I take an appropriate action by notifying my supervisors or completing paperwork. 1 2 3 4 5

When I see a vehicle that needs maintenance, I inform my supervisors or the appropriate person about it. 1 2 3 4 5

When I experience a near-miss event, I talk to people at work to learn from the incident. 1 2 3 4 5

Appendices 191

If I see my co-workers doing something risky while driving for work (e.g., using their mobile phone or speeding), I talk to them about the hazards of their risky behaviours. 1 2 3 4 5

I would intervene if I see a co-worker doing something unsafe while driving for work. 1 2 3 4 5

I would intervene to stop safety violations of other drivers. 1 2 3 4 5

I encourage my co-workers to think about the dangers of driving unsafely. 1 2 3 4 5

I inform other drivers that violations of safety procedures will have negative consequences. 1 2 3 4 5

I encourage other drivers to follow safe work driving procedures. 1 2 3 4 5

I encourage new drivers to follow safe working procedures. 1 2 3 4 5

I go out of my way to look out for the safety of other drivers. 1 2 3 4 5

I take action to protect other drivers from risky situations. 1 2 3 4 5

I speak up with new ideas or changes in work driving safety. 1 2 3 4 5

I speak up about work driving safety issues that affect work drivers. 1 2 3 4 5

I speak up about safety concerns during team meetings or toolbox talks. 1 2 3 4 5

I speak up on work driving safety matters even if others might disagree. 1 2 3 4 5

I communicate my views about work driving safety issue, even if others would disagree. 1 2 3 4 5

192 Appendices

Appendix E: Participant Information Sheet for the Main Survey

PARTICIPANT INFORMATION FOR QUT RESEARCH PROJECT – Main Survey –

Beyond Compliance:

The investigation of Proactive Safety Behaviours in the Work Driving Context

QUT Ethics Approval Number 1600000824

RESEARCH TEAM

Principal Researcher: Ms Klaire Somoray PhD student Associate Researchers: Dr Ioni Lewis Principal Supervisor

Centre for Accident Research and Road Safety – Queensland (CARRS-Q)

Professor Cameron Newton Associate Supervisor School of Management, QUT Business School

Queensland University of Technology (QUT) Dr Darren Wishart External Supervisor

DESCRIPTION This project is being undertaken as part of a PhD study investigating the proactive safety behaviours performed out by workers in organisations that operate vehicle fleets. The purpose of this project is to examine the proactive safety behaviours within the work-related driving setting. Proactive safety behaviours are behaviours that individuals perform to ensure and/or to improve their safety and their coworkers’ while driving for work. These behaviours are above and beyond the employees’ official job’s description and organisational policies and procedures. You are invited to participate in this project because: i) driving is an essential part of your job duties and ii) you drive for work at least once a week. Supervisors, managers and work health and safety officers who also deals with work drivers as part of their jobs are also invited to participate in this project. PARTICIPATION Participation will involve completing an anonymous survey with Likert scale answers that will take approximately 25 – 30 minutes of your time. Questions will include: I encourage other drivers to follow safe working procedures. I fix safety issues that relates to work driving even if it is not my responsibility. I communicate my views about work driving safety issue, even if others would disagree. Your participation in this project is entirely voluntary. If you agree to participate, you do not have to complete any question(s) you are uncomfortable answering. Your decision to participate or not participate will in no way impact upon your current or future relationship with QUT, CARRS-Q or your organisation. If you do agree to participate, you can withdraw from the project without comment or penalty. However, as the survey is anonymous once it has been submitted it will not be possible to withdraw.

Appendices 193

EXPECTED BENEFITS It is expected that this project may not directly benefit you, but it may benefit the general community. The overall aim of the project is to improve employees’ safety while driving for work by understanding the proactive safety behaviours that work drivers may engage in and how we can encourage these behaviours in the workplace. To recognise your contribution the research team is offering the chance to win one of ten $100 Coles/Myer gift vouchers for completing the survey. Please note the opening date for entries for the first survey is 27th November 2017, the closing date for entries is 27th of February 2018. The Terms and Conditions of the prize draw can be located at: https://survey.qut.edu.au/survey-data/67/67667/media/62/6254.pdf RISKS There are minimal risks associated beyond normal day-to-day living associated with your participation in this project. The questions on this survey will focus on positive and proactive behaviours in the workplace as well as questions about your organisations’ safety climate, safety leadership, motivation and risk perception. If there are items that you are not comfortable in answering, you have the option to skip them. All information you provide will be treated as confidential and will not be shared. Any reports, findings or papers produced using data collected during this research project will only contain aggregate data (i.e. participants grouped together) in order to ensure no individual or identifying details are included. QUT provides for limited free psychology, family therapy or counselling services (face-to-face only) for research participants of QUT projects who may experience discomfort or distress as a result of their participation in the research. Should you wish to access this service please call the Clinic Receptionist on 07 3138 0999 (Monday–Friday only 9am–5pm), QUT Psychology and Counselling Clinic, 44 Musk Avenue, Kelvin Grove, and indicate that you are a research participant. Alternatively, Lifeline provides access to online, phone or face-to-face support, call 13 11 14 for 24 hour telephone crisis support. PRIVACY AND CONFIDENTIALITY All comments and responses are anonymous and will be treated confidentially unless required by law. The names of individual persons are not required in any of the responses. Any data collected as part of this project will be stored securely as per QUT’s Management of research data policy. Please note that non-identifiable data from this project may be used as comparative data in future projects or stored on an open access database for secondary analysis. CONSENT TO PARTICIPATE The return of the completed survey via mail or online is accepted as an indication of your consent to participate in this project. QUESTIONS / FURTHER INFORMATION ABOUT THE PROJECT If you have any questions or require further information please contact one of the listed researchers. Klaire Somoray [email protected] 07 3138 4909 Ioni Lewis [email protected] 07 3138 4966

194 Appendices

Cameron Newton [email protected] 07 3138 2523 CONCERNS / COMPLAINTS REGARDING THE CONDUCT OF THE PROJECT QUT is committed to research integrity and the ethical conduct of research projects. However, if you do have any concerns or complaints about the ethical conduct of the project you may contact the QUT Research Ethics Advisory Team on 07 3138 5123 or email [email protected]. The QUT Research Ethics Advisory Team is not connected with the research project and can facilitate a resolution to your concern in an impartial manner. THANK YOU FOR HELPING WITH THIS RESEARCH PROJECT. PLEASE PRINT THIS SHEET FOR YOUR INFORMATION.

Appendices 195

Appendix F: Main Survey

PROACTIVE SAFETY BEHAVIOURS AND WORK-RELATED

DRIVING The Centre for Accident Research and Road Safety Queensland (CARRS-Q) at Queensland University of Technology (QUT) is currently doing a research project on work driving safety. A major component of the project is to understand the positive and proactive behaviours that work drivers and their supervisors perform in order to ensure employees' safety while driving for work. The questionnaire is strictly confidential and the research team will combine all the collected responses, therefore, your specific response will be anonymous. The questionnaire should only take 25 to 30 minutes to complete. THANK YOU FOR YOUR ASSISTANCE. If you would like to participate in the study, consent is required. Please indicate whether you are providing your consent to participate in this study:

Yes, I would like to provide my consent to participate in the study

No, I do not want to participate in the study

INSTRUCTIONS For each of the following questions, please select the answer which BEST reflects your views and/or experiences. Please indicate your answer by circling the best choice that corresponds with your opinions or by writing your answer in the space provided. Thank you.

196 Appendices

PLEASE NOTE: "Driving for work" includes driving for the purposes of work-related tasks (e.g., going to meetings, driving to see a client or a customer, driving to a work site for a job). This also INCLUDES commuting between work and home.

1. The following questions will ask about your work driving exposure. Please circle one number from 1 to 8: Approximately, how many hours PER WEEK do you normally drive for work?

1 – 10 hours ........................ 11 – 20 hours ...................... 21 – 30 hours ...................... 31 – 40 hours ...................... 41 – 50 hours ...................... 51 – 60 hours ...................... 61 hours or more ................. I do not drive for work ..........

1 2 3 4 5 6 7 8

Approximately, how many kilometres do you drive EACH YEAR for work?

1 – 10,000 kms .................... 10,001 - 20,000 kms ........... 20,001 - 30,000 kms ........... 30,001 - 40,000 kms ........... 40,001 - 50,000 kms ........... 50,001 - 60,000 kms ........... 60,001 kms or more ............ I do not drive for work ..........

1 2 3 4 5 6 7 8

If choose “I do not drive for work”, we thank you for your interest in the survey; however, at the present time, we are looking for participants who currently drive for work purposes. As such, we ask that you please do not continue with the survey. Thank you.

Appendices 197

2. The following questions will ask some information about you and your

work vehicle use

Are you? Male ......................................................................... Female .....................................................................

1 2

What is your age (in years)?

____________________ (years old)

How long have you held your driver’s licence (in years)?

____________________ (years)

How long have you been working for this organisation (in years)?

____________________ (years)

What type of vehicle do you mainly drive for work?

Car/ Sedan/Wagon .................................................. 4WD/SUV ................................................................ Heavy Vehicle .......................................................... Other (please specify____________________)

1 2

3

4

When driving for work, in a typical week, where do you do most of this driving?

Asphalt/Bitumen Roads ........................................... Dirt Roads ................................................................ Combination of Asphalt/Bitumen Roads and Dirt Roads ...................................................................... Other (please specify____________________)

1 2

3

4

5

198 Appendices

3. The following questions will ask about your exposure to work-related driving risks. Before answering the questions, please read the definitions below:

‘Driving for work’ - includes driving for the purposes of work-related

tasks. This INCLUDES commuting between work and home.

‘Crashes’ - any incident involving a motor vehicle that resulted in an injury or any damage to a vehicle, property or person (pedestrian).

'Near-miss' - any incident involving a motor vehicle where you narrowly avoided a collision with a vehicle, property or person (pedestrian).

‘Offences’ - any incident for which you were fined or incurred a loss of demerit points. This DOES NOT INCLUDE parking offences.

In the past 12 months, how many crashes have you experienced while driving for work?

None ....................................................................... One crash ............................................................... Two crashes ........................................................... Three or more crashes ...........................................

1 2

3

4

In the last 12 months, how many near-misses have you experienced while driving for work?

None ....................................................................... One near-miss ........................................................ Two near-misses .................................................... Three or more near-misses ....................................

1 2

3

4

In the last 12 months, how many traffic offences have you experienced while driving for work?

None ....................................................................... One traffic offence .................................................. Two traffic offences ................................................ Three or more traffic offences ................................

1 2

3

4

Appendices 199

4. The following questions will ask you about some behaviours that you may perform at work.

Please circle one number for each item from 1 = “Never” to 5 = “Always” N

ever

Rar

ely

So

met

imes

Mo

st o

f th

e

tim

e

Alw

ays

I speak up about safety concerns during team meetings or toolbox talks.

1 2 3 4 5

I speak up about work driving safety issues that affect work drivers even if it makes me unpopular.

1 2 3 4 5

I communicate my views about work driving safety issue, even if others would disagree.

1 2 3 4 5

I go out of my way to educate new employees about work driving safety procedures.

1 2 3 4 5

I go above my duties to help other drivers with their work driving safety responsibilities.

1 2 3 4 5

I go out of my way to help other drivers learn more about safe work driving practices.

1 2 3 4 5

I go above my duties to help my co-workers drive safely.

1 2 3 4 5

I actively seek feedback from my supervisor about my work driving.

1 2 3 4 5

I actively seek feedback from my co-workers and passengers about my work driving.

1 2 3 4 5

I ask people at work for feedback about my driving.

1 2 3 4 5

I encourage new drivers to follow safe working procedures.

1 2 3 4 5

I inform other drivers that violations of safety procedures will have negative consequences.

1 2 3 4 5

I encourage other drivers to follow safe working procedures.

1 2 3 4 5


1 2 3 4 5


1 2 3 4 5


1 2 3 4 5

When driving for work, I plan extra journey time and breaks for bad weather, traffic congestion, etc.

1 2 3 4 5

I fix safety issues that relate to work driving even if it is not my responsibility.

1 2 3 4 5

If I see something unsafe, I go out of my way to take care of it.

1 2 3 4 5

I go above my duties to resolve driving safety issues at work.

1 2 3 4 5

200 Appendices

If I notice a defect in the vehicle I am driving, I take an appropriate action by notifying my supervisors or completing the paperwork.

1 2 3 4 5

When I see a vehicle that needs maintenance, I inform my supervisors or the appropriate person about it.

1 2 3 4 5

I make sure that vehicle defects are attended to quickly.

1 2 3 4 5

I take action to protect other drivers from risky situations.

1 2 3 4 5

I go out of my way to look out for the safety of other drivers.

1 2 3 4 5

I look after the safety of other drivers in my organisation.

1 2 3 4 5

I would intervene if I saw a co-worker doing something unsafe while driving for work.

1 2 3 4 5

If I see my co-workers doing something risky while driving for work (e.g., using their mobile phone or speeding), I would intervene.

1 2 3 4 5

I would intervene to stop safety violations of other drivers.

1 2 3 4 5

I follow the work driving safety policies and procedures.

1 2 3 4 5

I attend mandatory meetings or toolbox talks for work driving safety.

1 2 3 4 5

I comply with the organisation’s rules and regulations on work driving safety.

1 2 3 4 5

I go to the required training and education programs on driving safety provided by my organisation. 1 2 3 4 5

Appendices 201

In your job role, do you supervise a work driver? Yes No If you answered YES, please complete the questions on the next page under the section named SUPERVISORS. If you answered NO, please stay on this page and answer the questions under the section named WORK DRIVERS.

WORK DRIVERS

5a. The following questions will ask about your relationship with your supervisors. When answering the following questions, think about the supervisor who is responsible for the daily management of your work driving (e.g., the one who you report to regarding your driving for work).

Please indicate how strongly do you agree or disagree with the following statements. S

tro

ng

Dis

agre

e

Dis

agre

e

Neu

tral

Ag

ree

Str

on

gly

Ag

ree

I usually know where I stand with my supervisor.

1 2 3 4 5

My supervisor understands my job problems and needs.

1 2 3 4 5

My supervisor recognises my potential.

1 2 3 4 5

Regardless of how much formal authority my supervisor has, he or she would personally help me solve problems in my work.

1 2 3 4 5

I can count on my supervisor to help me out even at his or her own expense, when I really need it.

1 2 3 4 5

My supervisor has enough confidence in me that he or she would defend and justify my decisions even if I am not present to do so.

1 2 3 4 5

My working relationship with my supervisor is effective.

1 2 3 4 5

202 Appendices

SUPERVISORS

5b. The following questions will ask about your relationship with work drivers that you supervise. When answering the following questions, think about the work drivers whom you are responsible for.


tro

ng

Dis

agre

e

Dis

agre

e

Neu

tral

Ag

ree

Str

on

gly

Ag

ree

The work drivers that I supervise usually know where they stand with me.

1 2 3 4 5

I understand the problems and needs of my work drivers.

1 2 3 4 5

I recognise the potential of my work drivers.

1 2 3 4 5

Regardless of how much formal authority I have, I would personally help my work drivers with their work problems.

1 2 3 4 5

I would use my formal authority to help my work drivers, even at my own expense.

1 2 3 4 5

I have enough confidence in my work drivers that I would support their decisions.

1 2 3 4 5

I have an effective relationship with the work drivers that I supervise.

1 2 3 4 5

Halfway there!

6. The following questions will ask you about your organisation's views on safety while driving for work. The first part will ask about your company as a whole and the next part will ask questions about the supervisors who you report to regarding your work driving.

Please indicate how often you experience the following situations. N

ever

Rar

ely

So

met

imes

Mo

st o

f th

e t

ime

Alw

ays

My company….

…reacts quickly to solve the problem when told about work driving safety concerns.

1 2 3 4 5

…is strict about driving safely even when work tasks fall behind schedule.

1 2 3 4 5

Appendices 203

…listens carefully to workers’ ideas about improving work driving safety.

1 2 3 4 5

...invests a lot in work driving safety training for workers.

1 2 3 4 5

…uses any available information to improve existing work driving safety rules.

1 2 3 4 5

…tries to continually improve driving safety levels in each department.

1 2 3 4 5

If you are a work driver, think about your direct supervisors (who are responsible for your work driving) and answer how often they do the following. If you are a supervisor for a work driver, think about the management-level and answer how often they do the following. N

ever

Rar

ely

So

met

imes

Mo

st o

f th

e t

ime

Alw

ays

My supervisors or the supervisors in my company…

...have discussions with employees on how to improve safety while driving for work.

1 2 3 4 5

…compliment employees who pay special attention to work driving safety.

1 2 3 4 5

…are strict about driving safely even when workers are tired or stressed.

1 2 3 4 5

…frequently talk about work driving safety issues throughout the work week.

1 2 3 4 5

…are strict with driving safety rules even when work falls behind schedule.

1 2 3 4 5

…uses explanations (not just compliance) to get employees to act safely while driving for work.

1 2 3 4 5

7. The following questions will ask you about your general feelings towards safety while driving for work.


tro

ng

Dis

agre

e

Dis

agre

e

Neu

tral

Ag

ree

Str

on

gly

Ag

ree

I anticipate risks or safety problems when driving for work, thinking of the possible alternative scenarios if problems arise.

1 2 3 4 5

I look at work driving safety issues from different perspectives to find appropriate solutions.

1 2 3 4 5

204 Appendices


tro

ng

Dis

agre

e

Dis

agre

e

Neu

tral

Ag

ree

Str

on

gly

Ag

ree

Even before crashes happen, I think about various risky situations that may compromise safety when driving for work.

1 2 3 4 5

I look ahead to ensure that safety issues relating to driving for work are addressed.

1 2 3 4 5

I recognise that I am able to make significant contributions to the work driving safety within my department.

1 2 3 4 5

I recognise that I have an influence on matters that relate to work driving safety.

1 2 3 4 5

I recognise that my own actions have great importance for the driving safety of the work team.

1 2 3 4 5

I recognise that most of the safety problems that occurs while driving for work are under one’s own control.

1 2 3 4 5

I strive hard to be an example of someone who has strong commitment to work driving safety.

1 2 3 4 5

I feel that it's my responsibility to discuss work driving safety issues with my co-workers.

1 2 3 4 5

I feel responsible for taking the initiative in suggesting improvements to workers' safety while driving for work.

1 2 3 4 5

I depend on me to make improvements to my organisation's work driving safety.

1 2 3 4 5

I follow the work driving safety rules and regulations set by my organisation to avoid getting in trouble.

1 2 3 4 5

I try to avoid making mistakes or violating rules when driving for work because I do not want to get in trouble.

1 2 3 4 5

I follow the rules when driving for work because I do not want to get caught.

1 2 3 4 5

Appendices 205

Appendix G: Collection Method - Online versus Hardcopies

Online (n =

219) Hardcopy (n =

75)

BCa 95% Confidence

Interval M SD M SD t df p-value Lower Upper

Educate 14.36 5.79 13.27 5.07 1.67 142.16 0.107 -0.33 2.52

Voice 8.84 3.60 9.52 3.39 -1.31 289.00 0.181 -1.55 0.33

Fixing Safety

Issues 13.43 4.46 13.71 3.54 -0.53 156.95 0.617 -1.26 0.78

Vehicle

Maintenance 11.88 3.60 13.11 1.91 -3.67 238.76 0.001 -1.84 -0.55

Feedback Inquiry 5.73 3.29 4.81 2.25 2.98 191.52 0.003 0.33 1.64

Intervene 10.65 3.18 11.47 2.53 -1.91 289.00 0.037 -1.51 -0.04

Notes. N = 294; Analysis was conducted using independent sample t-test; df = degrees of freedom; BCa = Bootstrap Confidence Intervals based on 10,000 samples; highlighted cells indicate statistically significant differences based on exclusion of zero between the lower and upper intervals

206 Appendices

Appendix H: Gender differences (male versus female) on proactive safety

behaviours

Male Female

BCa 95% Confidence


Educate 14.71 5.67 13.48 5.54 1.92 291.00 0.041 -0.09 2.57

Voice 9.06 3.46 8.97 3.66 0.28 291.00 0.755 -0.69 0.88

Fixing Safety

Issues 13.52 4.31 13.48 4.18 0.10 291.00 0.914 -0.86 0.94

Vehicle

Maintenance 12.41 3.28 11.98 3.30 1.17 291.00 0.202 -0.33 1.16

Feedback

Inquiry 5.92 3.30 5.09 2.80 2.43 279.41 0.011 0.11 1.51

Intervene 10.77 3.13 10.95 2.97 -0.47 291.00 0.623 -0.78 0.44


Appendices 207

Appendix I: Supervisory role differences on proactive safety behaviours

Has a supervisor

role (n = 31) No supervisor role

(n = 263)

BCa 95% Confidence


Educate 17.58 5.41 13.67 5.52 3.74 292.00 0.001 1.73 6.00

Voice 11.10 3.11 8.76 3.53 3.52 292.00 0.001 1.16 3.44

Fixing Safety

Issues 15.45 3.95 13.27 4.22 2.74 292.00 0.002 0.64 3.68

Vehicle

Maintenance 13.42 2.78 12.05 3.33 2.21 292.00 0.014 0.10 2.43

Feedback Inquiry 7.23 3.80 5.29 2.93 3.36 292.00 0.009 0.55 3.39

Intervene 12.81 2.26 10.63 3.05 4.87 44.04 0.001 1.25 3.04


208 Appendices

Appendix J: Differences on proactive safety behaviour scores based on the

frequently used road type when driving for work

Asphalt / Bitumen roads

(n = 249) Combination

(n = 39)

BCa 95% Confidence

Intervals

M SD M SD t df p-value Lower Upper

Educate 13.68 5.60 16.28 5.17 4.74 286 0.006 -4.42 -0.79

Voice 8.71 3.57 10.74 2.94 6.03 286 0.001 -3.00 -0.97

Fixing Safety

Issues 13.30 4.29 14.67 3.86 1.79 286 0.036 -2.70 -0.10

Vehicle

Maintenance 12.03 3.44 13.26 1.98 3.02 79.51 0.002 -2.03 -0.46

Feedback

Inquiry 5.30 2.97 6.41 3.41 5.61 286 0.051 -2.33 -0.01

Intervene 10.84 3.00 11.18 3.26 0.26 286 0.550 -1.41 0.83


Appendices 209

Appendix K: Differences on proactive safety behaviour scores based on the

frequently used vehicle when driving for work

Car / Sedan /

Wagon (n = 178) SUV / 4WD

(n = 79) Heavy Vehicle

(n = 20)

M SD M SD M SD F df p-value

Educate 13.49 5.74 14.38 5.66 16.15 3.48 2.69 3 0.05

Voice 8.26 3.62 9.77 3.24 10.55 2.35 8.79 3 0.00

Fixing Safety Issues 13.08 4.39 14.03 4.05 14.80 3.58 1.68 3 0.17

Vehicle Maintenance 11.53 3.71 13.04 2.24 13.25 2.45 6.95 3 0.00

Feedback Inquiry 5.48 3.19 5.29 2.77 5.70 2.89 0.66 3 0.58

Intervene 10.74 3.19 11.18 2.91 10.50 3.05 0.53 3 0.67

Notes. N = 294; Analysis is conducted using one-way ANOVA; df = degrees of freedom; highlighted cells indicate statistically significant differences between groups

210 Appendices