t emporal and s patial variability of the determinants of satisfaction with public transport in...
Post on 01-Apr-2015
217 Views
Preview:
TRANSCRIPT
TEMPORAL AND SPATIAL VARIABILITY OF THE DETERMINANTS OF SATISFACTION WITH PUBLIC TRANSPORT IN SWEDEN
MASTER THESIS
TRANSPORT AND GEOINFORMATION
TECHNOLOGIES
FERNÁNDEZ ABENOZA, Roberto
Supervisors: Yusak O. Susilo
Oded Cats Division of Transport and Location analysisTransport Science DepartmentRoyal Institute of Technology, Stockholm
PROBLEM DESCRIPTION
OBJECTIVE
SCOPE AND LIMITATION
METHODOLOGY
RESULTS DESCRIPTIVE STATSC. CORRELATIONSFACTOR ANALYSISREGRESSION M.
PRIORITY MAPS
CONCLUSIONS & RECOMMENDATIONS
BackgroundWorld trends: Increase population, larger need of mobility and accessibility, longer trips, road safety, environmental problems, etc. PTx2 Strategy (2009-2025 increase use transport by 100%)
Europe: EC White Paper 2011 sets Tranport sector goals by 2050
Sweden issues: Need for socially and environmental sustainable society (transport act 2009); Swedish doubling project (PT modal share up to 24% by 2020); Increase of contribution/subsidies to landsting
Increasing Satisfaction with Public Transport (PT) service very important to increase ridership.
Important to razionalize costs while maximising it benefits.
PROBLEM DESCRIPTION
OBJECTIVE
SCOPE AND LIMITATION
METHODOLOGY
RESULTS DESCRIPTIVE STATSC. CORRELATIONSFACTOR ANALYSISREGRESSION M.
PRIORITY MAPS
CONCLUSIONS & RECOMMENDATIONS
Customer satisfaction• Overall fullfilment with customer’s expectations (Tyrinopoulos and Antoniou,
2008)
• Outcome of cumulative satisfaction through time or single-experiences, or the two. Negative Single encounters: NCI (Friman et al, 2001; Friman and Fellesson, 2009; Redman et al, 2013)
• Increase of customer satisfaction leads to a growth in customer loyalty (Barabino et al, 2012)
Adaptation of ECSI customer satisfaction model
PROBLEM DESCRIPTION
OBJECTIVE
SCOPE AND LIMITATION
METHODOLOGY
RESULTS D.STATS
C. CORRELATIONSFACTOR ANALYSISREGRESSION M.
PRIORITY MAPS
CONCLUSIONS & RECOMMENDATIONS
Purpose
Which are the most important QoSA that contribute to customer satisfaction with PT services in Sweden?
andAre they consistent over time and space?
- Overview of drivers of satisfaction with PT services focusing on supply characteristics of the service
- Knowledge whether the determinants of satisfaction remain unchanged over time. Detect future trends and stability/variability of certain determinants (QoSA)
- Insight into the uniformity or nonuniformity of the determinants across geographical areas
- Ascertain and Identify priority areas for different regions and years, in terms of importance, current performance and trend.
PROBLEM DESCRIPTION
OBJECTIVE
SCOPE AND LIMITATION
METHODOLOGY
RESULTS DESCRIPTIVE STATSC. CORRELATIONSFACTOR ANALYSISREGRESSION M.
PRIORITY MAPS
CONCLUSIONS & RECOMMENDATIONS
Groups interested in outcome
Why is it important?
It is of capital importance for service performance monitoring, market analysis, benchmarking and the identification of priority areas
For whom?
Study is instrumental in supporting PT providers, Regional PT authorities as well as County councils and Municipalities to provide a PT service to match provision with their user’s needs and thus foster customer satisfaction and ultimately increase PT ridership
PROBLEM DESCRIPTION
OBJECTIVE
SCOPE AND LIMITATION
METHODOLOGY
RESULTS DESCRIPTIVE STATSC. CORRELATIONSFACTOR ANALYSISREGRESSION M.
PRIORITY MAPS
CONCLUSIONS & RECOMMENDATIONS
Scope and limitation
LIMITATIONS
A. Focus on PT users users keeping them happy and increasing ridership . No investigation of only car users
B. Variability over space limited to 5 different county regions levels.
Geographical Scope
Time frame = 2001-2013
PROBLEM DESCRIPTION
OBJECTIVE
SCOPE AND LIMITATION
METHODOLOGY
RESULTS DESCRIPTIVE STATSC. CORRELATIONSFACTOR ANALYSISREGRESSION M.
PRIORITY MAPS
CONCLUSIONS & RECOMMENDATIONS
Work-flow
Diagram
PROBLEM DESCRIPTION
OBJECTIVE
SCOPE AND LIMITATION
METHODOLOGY
RESULTS D.STATS
C. CORRELATIONSFACTOR ANALYSISREGRESSION M.
PRIORITY MAPS
CONCLUSIONS & RECOMMENDATIONS
Quality of Service Attributes (QoSA)
SERVICE QUALITY
AREAS QUALITY OF SERVICE ATTRIBUTES
Time Length of Trip time Adherence to schedule
Information General information Information on planned
changes
Information on unplanned
changes Accessibility Accessibility Ticket Easiness of travel
Availability Operation Network Availability
Network
Comfort On-board
conditions Ride
Comfort
Ambient off-board
conditions
Use of Travel Time
Traffic Stress
avoidance Customer care Staff and assistance Customer interface
Security Freedom from crime Environmental
impact Environmentally friendly
PROBLEM DESCRIPTION
OBJECTIVE
SCOPE AND LIMITATION
METHODOLOGY
RESULTS DESCRIPTIVE STATSC. CORRELATIONSFACTOR ANALYSISREGRESSION M.
PRIORITY MAPS
CONCLUSIONS & RECOMMENDATIONS
Application of Weights
Map of Sweden per County
Cartogram regarding Population per county
in 2001
Cartogram regarding sample size from dataset in 2001
Weighted by county, urban area (Tätorter) and age
PROBLEM DESCRIPTION
OBJECTIVE
SCOPE AND LIMITATION
METHODOLOGY
RESULTS DESCRIPTIVE STATSC. CORRELATIONSFACTOR ANALYSISREGRESSION M.
PRIORITY MAPS
CONCLUSIONS & RECOMMENDATIONS
Socio-demographic & mobility profile of the weighted sample
Gender
Age groups
Occupation
Disability (2001,2007-2013)
Distance work/school
Urban Area size(in thousands)
Density Population(hab/km2)
Car availability
Frequency travel car(2010-2013)
Frequency travel by PT
Most used PT mode(2011-2013)
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Male Female
15-24 25-34 35-44 45-54 55-64 65-75
Retired / Sick leave Student
Yes No
<1km 1.0 - 3.0 km 3.1 - 5.0 km 5.1 - 10.0 km 10.1 - 30.0 km 30.1 - 50.0 km
>50km
Stockholm-Göteborg-Malmö 100-150
75-100
50-75 <50
<10 10,1-21 21,1-41 41,1-56 51,1-116 >116
Yes No
Daily Weekly Monthly More rarely Never
Daily Weekly Monthly More rarely
City bus Regional bus Commuter train Metro Regional
trainTram
Employed / Self-employed / Employee
PROBLEM DESCRIPTION
OBJECTIVE
SCOPE AND LIMITATION
METHODOLOGY
RESULTS DESCRIPTIVE.STATSC. CORRELATIONSFACTOR ANALYSISREGRESSION M.
PRIORITY MAPS
CONCLUSIONS & RECOMMENDATIONS
Overall Satisfaction & QoSA over time
PROBLEM DESCRIPTION
OBJECTIVE
SCOPE AND LIMITATION
METHODOLOGY
RESULTS DESCRIPTIVE STATSC. CORRELATIONSFACTOR ANALYSISREGRESSION M.
PRIORITY MAPS
CONCLUSIONS & RECOMMENDATIONS
Cross-correlations• Intercorrelations between QoSA and Overall Satisfaction all positive: Halo
Effect• Highest intercorrelation between Quality and Popularity and Overall
Satisfaction. Explains why higher Overall Satisfaction in Medium/Small counties.
• Strong intercorrelation between Frequency of travel by PT and Relevance implies that a useful and functional PT service favours to stimulate ridership.
• Very high intercorrelation between Loyalty and Overall Satisfaction in consonance with that high OS leads to higher loyalty
PROBLEM DESCRIPTION
OBJECTIVE
SCOPE AND LIMITATION
METHODOLOGY
RESULTS DESCRIPTIVE STATSC. CORRELATIONS
FACTOR ANALYSISREGRESSION M.
PRIORITY MAPS
CONCLUSIONS & RECOMMENDATIONS
Factor Analysis (PCA)
The component solution explains a total of 57.3% of the variance.
Soft component contributing a 37.5% and Functional and Information components explaining 11.3% and 8.5%, respectively.
PROBLEM DESCRIPTION
OBJECTIVE
SCOPE AND LIMITATION
METHODOLOGY
RESULTS DESCRIPTIVE.STATSC. CORRELATIONSFACTOR ANALYSISREGRESSION M.
PRIORITY MAPS
CONCLUSIONS & RECOMMENDATIONS
Models specification
MODELS
Explanatory variables included Time and Geographical frame
3 factors
from FA
12 QoSA
3 Frequency
PT dummy
12 Year
dummy
4 County dummy
Entire database
Per yearPer
County region
Joint Factors x x x x x
Joint QoSA x x x x x
Year-specific x x x x
Region-specific x x x x
• None of the socio-demographic and travel habits variable systematically significant over time and space. Low explanatory power, thus removed
• No interaction variables were found
PROBLEM DESCRIPTION
OBJECTIVE
SCOPE AND LIMITATION
METHODOLOGY
RESULTS DESCRIPTIVE STATSC. CORRELATIONSFACTOR ANALYSISREGRESSION M.
PRIORITY MAPS
CONCLUSIONS & RECOMMENDATIONS
S.C. t-stats S.C. t-stats
(Constant) ,474 42,012 3,544
Soft Factor ,324 178,455
Ride comfort ,020 10,855
On-board conditions ,044 23,515
Staff and assistance ,083 45,277
Freedom from crime ,136 70,470
Functional Factor ,304 163,500
Network ,099 50,171
Operation ,165 84,950
Length of Trip time ,156 79,551
Information Factor ,278 148,678
Accessibility Ticket ,043 25,078
Information on planned changes
,044 22,206
Information on unplanned changes
,066 34,006
Customer interface ,183 99,119
General information ,047 27,758
Daily ,042 22,342 ,035
One Few time week ,054 29,563 ,059
OneFew time month ,035 20,085 ,048
Västra Götalands ,024 13,284 ,018 9,639
Skåne -,007 -3,485
Medium size ,016 8,04 ,006 3,008
Small size ,034 17,063 ,022 11,115
2002 0,005* 2,048
2003 ,007 2,723 ,010 3,478
2004 ,016 5,814 ,017 6,163
2005 ,011 4,172 ,011 3,798
2006 ,008 2,903 ,010 3,494
2007 ,016 5,232 0,006* 2,003
2008 ,018 6,039 0,007* 2,323
2009 ,017 5,701 0,005** 1,742
2010 ,015 4,939 0,008* 2,347
2011 (-)0,006** -1,887
2012 -,013 -4,045 -,023 -6,756
2013 -,014 -4,250
R-square
Adjusted R-square
Number of observations 397861 397861
Joint QoSA Joint Factors
,502 ,505,502 ,505
Joint ModelsJoint Factors Model
Factors have similar influenceSequence: Soft, Functional, Information
Joint QoSA Model
Customer interface
Operation
Length of trip time
Freedom from crime
Ride Comfort
Joint contribution of QoSA into factors.
Largest differences. Sequence: Functional, Information, Soft
* Sig. 0.05 ** Sig. 0.1 Otherwise Sig. 0.01
PROBLEM DESCRIPTION
OBJECTIVE
SCOPE AND LIMITATION
METHODOLOGY
RESULTS DESCRIPTIVE STATSC. CORRELATIONSFACTOR ANALYSISREGRESSION M.
PRIORITY MAPS
CONCLUSIONS & RECOMMENDATIONS
Priority map per Year (2001-2013)
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18 0.20 0.22
0.70
0.75
0.80
0.85
0.90
0.95
1.00
1.05
1.10
1.15
1.20
1.25
1.19213236263961
1.0286275728068
0.9331786418753030.923416448979771
1.05503612548912
1.09216371118337
1.06885907068224
0.982606055600007
1.11281152974792
0.899928488940797
0.780314245497073
0.830556137558788
1.19616939840976
1.04649714733439
0.9378109514248190.922581683797955
1.05190567270466
1.09240909521426
1.06610943940248
0.986073363245494
1.11366574224464
0.891296441316355
0.758909224570582
0.832671460207872
1.188714938239
1.03680475213205
0.9299340758978690.920372313516002
1.0391093128259
1.088859526192041.06194869606625
0.989672429859854
1.11678514424096
0.903876470943654
0.76669863352118
0.848590958554591
1.1731082110579
1.03468743179023
0.9212663115619480.909632534906267
1.04539827904512
1.09176244337643
1.06348716340652
0.98655124280051
1.10730582482253
0.927637852232024
0.782487479315381
0.852386325032001
1.159440637599
1.06430430901602
0.9134698003523110.897016994166968
1.0543610155894 1.09393584832375
1.06123167459492
0.977759319459731
1.097746027727
0.93563697285288
0.783482363928668
0.852941692097737
1.16201260507429
1.09188605219324
0.9091586962845720.897905691534066
1.05381004010091
1.09185974578491
1.06069556720771
0.978196305073195
1.09006004619455
0.938890261510303
0.776576471973253
0.859209721327449
1.15021905266384
1.12528800799704
0.903424507869520.893072979867936
1.04939726525025
1.09487988289315
1.05593737454196
0.973486720133154
1.08643658063261
0.93744212512532
0.782115039575069
0.85444429459576
1.105323528176011.11600542637411
0.9209957413058150.932539963749078
1.03581603547094
1.08026582866986
1.04222991830668
0.996010619998903
1.08692684267162
0.932656573742018
0.789739182627867
0.852078415220661
1.11095963937683
1.13754319527372
0.9240781436634250.936296886127767
1.03396489878186
1.07790134189942
1.037156549893230.996377143812037
1.09156327459066
0.927970713966176
0.77645059465367
0.840506211116248
1.097
1.137
0.9310.946
1.044
1.080
1.040
0.994
1.097
0.923
0.766
0.839
1.078
1.145
0.9300.948
1.052
1.083
1.044
0.994
1.100
0.921
0.758
0.837
1.064
1.141
0.931
0.952
1.049
1.081
1.040
0.995
1.097
0.931
0.774
0.843
1.062
1.145
0.923
0.953
1.046
1.080
1.051
1.000
1.101
0.933
0.768
0.835
Customer's Priorities
Sati
sfac
tion
inde
x: 1
Ave
rage
Sas
tisf
acti
on S
core
=3,5
2Information on unplanned changes
Customer interface
Freedom from crime
Information on planned changes
Network
General informa-tion
Staff and assistance
Operation
Length of trip time
Ride Comfort
Accessibility ticket
On-board conditions
Possible overkill
Lower Priority
PRIORITY AREA
Keep up the good
work
1 colour per Quality of Service Attribute. Ex. Red for Network 1=2001; 2=2002…13=2013
PROBLEM DESCRIPTION
OBJECTIVE
SCOPE AND LIMITATION
METHODOLOGY
RESULTS DESCRIPTIVE STATSC. CORRELATIONSFACTOR ANALYSISREGRESSION M.
PRIORITY MAPS
CONCLUSIONS & RECOMMENDATIONS
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18 0.20 0.22
0.70
0.75
0.80
0.85
0.90
0.95
1.00
1.05
1.10
1.15
1.20
1.25
1.177
1.141
1.035
0.979
0.970
1.065
1.029 0.993
1.065
Travel Information on planned changes
Travel Information on unplanned changes
0.804
Stockholms Västra Götalands Skåne Medium size Small size
Customer's Priorities
Satis
facti
on in
dex:
1
Aver
age
Sasti
sfac
tion
Scor
e=3,
53
PRIORITY AREA
Possible overkill
Lower Priority
Keep up the good
work
Priority map per County-region
PROBLEM DESCRIPTION
OBJECTIVE
SCOPE AND LIMITATION
METHODOLOGY
RESULTS DESCRIPTIVE STATSC. CORRELATIONSFACTOR ANALYSISREGRESSION M.
PRIORITY MAPS
CONCLUSIONS & RECOMMENDATIONS
ConclusionsDeterminants of satisfaction with PT services in Sweden : Customer Interface, Freedom from Crime, Length of Trip Time, Operation, Network
Do the determinants of satisfaction remain unchanged over time?
Overall, the importance of the determinants of Satisfaction with PT service varies over time as can be observed by the horizontal movements of the QoSA The QoSA perdure in the same quadrants.
Do the determinants remain unchanged across regions in Sweden?
Yes, in general. Some QoSA that behave differently depending on the county-region, likewise Stockholm and Skåne for Network and Length of Trip Time.
How can service providers identify priority areas? Priority areas can be identified by adopting priority map which identify importance-priority areas.
PROBLEM DESCRIPTION
OBJECTIVE
SCOPE AND LIMITATION
METHODOLOGY
RESULTS DESCRIPTIVE STATSC. CORRELATIONSFACTOR ANALYSISREGRESSION M.
PRIORITY MAPS
CONCLUSIONS & RECOMMENDATIONS
RecommendationsGeneral• Customer satisfaction surveys: Maintain same set of QoSA over time
Increase key QoSAPrioritization of Customer Interface (Below satisfaction threshold)
Length of trip time (Not in Skåne)
Operation and Network (Not in Stockholm)
Keep security (Freedom from crime)
Watch Information on Planned changes
Information on Unplanned Changes. Might be dangerous.
Ticket Accessibility: Decreasing Satisfaction. Study why.
PROBLEM DESCRIPTION
OBJECTIVE
SCOPE AND LIMITATION
METHODOLOGY
RESULTS DESCRIPTIVE STATSC. CORRELATIONSFACTOR ANALYSISREGRESSION M.
PRIORITY MAPS
CONCLUSIONS & RECOMMENDATIONS
QoSATransport Operators
PT Regulators
Customer Interface XX
Operation X X
Length of trip time X X
Network X X
Freedom from Crime X X
Staff and Assistance XX
Information on unplanned changes XX
Information on Planned Changes XX
On-board Conditions XX
Ticket Accessibility XX
General Information XX
Ride Comfort XX
Degree of responsibility with the increase in satisfaction of the QoSA
PROBLEM DESCRIPTION
OBJECTIVE
SCOPE AND LIMITATION
METHODOLOGY
RESULTS DESCRIPTIVE STATSC. CORRELATIONSFACTOR ANALYSISREGRESSION M.
PRIORITY MAPS
CONCLUSIONS & RECOMMENDATIONS
Future research
1. Investigation of different geographical areas:- 25 Swedish Concession areas- Generate another groups of homogeneus regions
2. Inclusion of other QoSA - Application of a shorter time frame:
3. Investigation of determinants for different target groups
4. Analysis of seasonal variations (only available from 2010)
5. Analysis of Panel data (data not available)
PROBLEM DESCRIPTION
OBJECTIVE
SCOPE AND LIMITATION
METHODOLOGY
RESULTS DESCRIPTIVE STATSC. CORRELATIONSFACTOR ANALYSISREGRESSION M.
PRIORITY MAPS
CONCLUSIONS & RECOMMENDATIONS
Thanks for your attention!
Any questions?
rfa@kth.se
top related