Impact of brt implementation on commuter behavior in montorcycle dependent city case study ha noi city

In that situation, after success of BRT Bus Rapid Transit in Bogota and Curitiba and some China city, BRT which has high quality and outstandingly it has cost effective than other mass t

1

VIETNAM NATIONAL UNIVERSITY, HANOI

VIETNAM JAPAN UNIVERSITY

(All cap, bold, font size 14, Times New Roman)

FULL NAME OF THESIS AUTHOR

(All cap, bold, font size 14, Times New Roman)

Associate Prof Dr NGUYEN VAN A

(All cap, bold, font size 14, Times New Roman)

Hanoi, 20

(Regular, bold, font size 14, Times New Roman)

VIETNAM NATIONAL UNIVERSITY, HANOI

VIETNAM JAPAN UNIVERSITY

LUU DUY

IMPACT OF BRT IMPLEMENTATION ON

COMMUTER BEHAVIOR IN MOTORCYCLE

DEPENDENT CITY CASE STUDY: HA NOI CITY

MASTER’S THESIS

Hanoi, 2018

VIETNAM NATIONAL UNIVERSITY, HANOI

VIETNAM JAPAN UNIVERSITY

LUU DUY

IMPACT OF BRT IMPLEMENTATION ON COMMUTER BEHAVIOR IN MOTORCYCLE

DEPENDENT CITY CASE STUDY: HA NOI CITY

MAJOR: INFRACTRUCTER ENGINEER

CODE: ………

RESEARCH SUPERVISOR:

Ph.D NGUYEN HOANG TUNG

DR NGUYEN TIEN DUNG

Hanoi, 2018

3

Acknowledgment

In order to complete my master thesis with topic “Impact of BRT implementation to commuter behavior in Motorcycle dependence city”, I have received the support and assistance of my supervisors, VJU teachers, classmates and friends With deep affection, sincerity, allow me to express my profound gratitude to all my teachers and friends who have helped in the process of learning and completing my master thesis which would not be able to complete well without their help First of all, I would like to send to my supervisors Ph.D Nguyen Hoang Tung and Dr Nguyen Tien Dung, my respectful, deep thankfulness for their care and guidance that helped me during my hard time after internship I would like to send my sincere thanks to all the teachers of the Viet Nam Japan University who taught and imparted precious knowledge to me during the study period in university

Especially, I deeply appreciate my deep gratitude to Professor Kato, University of Tokyo, and Prof Nguyen Dinh Duc, Hanoi National University and Dr Phan Le Binh, Lecturer, JICA Long-term experts at VJU for their attention and kindly support unconditionally Therefore, I would like to express my deepest gratitude and special thanks

to my teachers

One can not fail to mention the support of my classmates in MIE – VJU, assistant of

my class Bui Hoang Tan and friends has also help me to finish the survey and thesis if there was no help of them, I would not finish thesis well Hence, I would like to express my gratitude to them for their help

Later, I would like to give my special thanks to my family, my lover who have given all support for me during my thesis

Given the limited time and experience of the thesis itself, these shortcomings can not

be avoided I hope to receive the advice and comments of the teachers to have improvement for my studying and working in my life

Thank you sincerely

Luu Duy

INTRODUCTION

CHAPTER 1I: LITERATURE REVIEW 4

1.1 What is Motorcycle dependence cities (MDCs) 4

1.2 Public transprot and people behavior in MDCs 4

CHAPTER III: PROBLEM STATEMENT 5

3.1 Problem statement 5

3.2 Study question 5

CHAPTER IV: RESEARCH OBJECTIVE AND SPECIFIC OBJECTIVES 6

4.1 Research Ojective 6

4.2 Research Specific Objectives 6

CHAPTER V: HYPOTHESIS 6

CHAPTER VI: APPOARCHES AND METHODOLOGY 9

6.1 Methodology 9

6.2 Sampling design 10

6.3 Questionair design 10

CHAPTER VII: CASE STUDY 11

7.1 Context background 11

7.2 Data collection 13

7.3 Analysis 15

7.3 Discucssion 35

CHAPTER VIII: CONCLUTSION 36

References 37

Appendix: The survey questionnaire 38

socio-In order to improve living standard and aim to build sustainable city, the government in that cities has planned the mass transit transport system (metro, Light Rail Transit) to solve the problem but it require high capital cost and long term construction while the cities still poor and limited resources In that situation, after success of BRT (Bus Rapid Transit) in Bogota and Curitiba and some China city, BRT which has high quality and outstandingly it has cost effective than other mass transit system (proved by Bangkok BRT system investment cost estimation in 2004 that show the cost is approximately 17 and 75 times cheaper than those

of Light Rail Transit (LRT) and Subway respectively (Thaned Satienam, 2006)), the city manager is recommended to be use BRT to solve the transport problems, improve the public transport system and especially change people awareness and behavior on using their commuting mode which will be the key factor for reducing the private vehicle in general and motorcycle in particular However, it still seems difficult to introduce BRT to these cities because its characteristic (planning, infrastructure ) and people behavior that likely stick on private vehicles due its advantages to fit their daily demand

Purpose of study

The main purpose of this study is the impact evaluation of BRT introduction to the people especially commuter who is the main group that contribute to daily traffic in motorcycle dependence cities It is expected the results will be helpful for the city managers

to evaluate the effectiveness of BRT introduction to their city for enhancing the transport system and improve the traffic conditions

Structure of the thesis

The thesis includes 6 chapters:

Chapter 1: Literature review

Chapter 2: Problems statement

Chapter 3: Research Objectives

Chapter 4: Hypothesis

Chapter 5: Approaches and Methodology

Chapter 6: Case study

Chapter 7: Conclusion and Recommendation

I: LITERATURE REVIEW

1.1 What is the Motorcycle dependence city (MDC):

Motorcycle dependence cities (MDCs) are the cities such as Ha Noi, Ho Chi Minh, Taipei that have special characteristic on land use and infrastructure that causing the domination of motorcycle in daily life The “motorcycle city” was firstly called by Barter (1999) to describe the situation of cities in Indonesia and Vietnam Futher, Khuat Viet Hung (2006) defined the MDCs by three indicators: (1) high motorcycle ownership (more than

350 motorcycles per 1000 inhabitants), (2) lack of public transport alternatives (less than one bus per 1000 inhabitants) and inadequate non-motorized trips compared with motorcycle, (3) high share of motorcycles in the traffic (more than 50%) and low modal split to public transport (lower than 20%) The people in MDCs use motorcycle for almost every activity in their daily life, meanwhile the public transport is facing a declining trend

of mode share in traffic due to lack of awareness for perceived quality and missing quality management (Vu Anh Tuan 2016)

1.2 Public transport and people behavior in MDCs:

Public transport is facing declining trend due to a lack of awareness for perceived quality and the missing quality management (Vu Anh Tuan 2016) Beside the decreasing of public transport, because the characteristic of MDCs that influence land-use pattern by allowing for the development of many areas which are only accessible by motorcycle, the people in MDCs use motorcycle to fit their demands, saving time and convenient and full-fill their trip which has high number in trip chaining than developed countries (average 4

7

trips/ day compare with average 2.5 trips/day due to poor land use control) (Vu Anh Tuan, 2015) The consequence of motorcycle domination is congestion and pollution but people in this MDCs not keen to use public buses, low intention to switch from their „motor biking‟ lifestyle even they are living in polluted environment (Poh Ng yen and Phung Thanh Phuong, 2015) Apart from the reason to less motivate on using bus is the private vehicle user usually has the negative look about public transport (Fujii and partner 2001) and in fact, the quality of bus system in these cities are low and underdeveloped, major group of people using bus are pupil, student and unemployment/ retirement (low income group) due

to cost saving, working group from 25 to 60 years old (middle and high income group), motorcycles take a dominant role for saving time and convenient (Cam Van, Vu Anh Tuan 2013; Vu Anh Tuan, 2012) Meanwhile, statistical data show an decease in walking rates throughout the world and in Asian countries (Ebru Cubukcu, 2013) and with the dependence of motorcycle and polluted environment, the people in MDCs do not have the positive awareness toward walking and the acceptable walking distance to use bus is within 400m from the station

CHAPTER II: PROBLEM STATEMENT:

2.1 Problem statement

Some researches stated that it is difficult to change the behavior of people in MDCs with BRT implementation then the city should focus on railway system development (Vu Anh Tuan, 2012) however the above conclusions may be uncertain because:

- The researchers evaluated the impact of BRT with Stated Preference method when BRT is not implemented (Vu Anh Tuan, 2012,2013,2015; Thanned Satiennam 2013, 2015; Nguyen Thi Cam Van, 2013)

- These studies do not focus on specific group, the main commuter group which impact to daily traffic is not well research

- Most of all, the behavior change on is not careful investigate in case of before and after implementation, for example the walking behavior which is associated with public transit and moving toward increased physical activities (Amy L Freeland, PhD, Shailendra

N Banerjee, PhD, Andrew L Dannenberg, MD, MPH, and Arthur M Wendel, 2013) and the basis of sustainable city to reduce pollution and healthy life(Rustam Khairi Zaharib, 2013)

Therefore, maybe the research will have different results with the data is collected in case of after BRT implementation and there are more aspects like walking behavior, perception that effect by the introduction

3.2 Research specific objectives:

The change on commuter behavior will be researched in 5 specific objectives: characteristics of BRT commuter, modal shift, perception, walking behavior, and trip chaining:

- Characteristic of BRT commuter: The change on characteristic of BRT commuter

compare with Bus commuter

- Modal shift: The change on weight of attributes on travel time and travel cost

between BRT commuter and bus commuter

- Commuter Perception: The change on perceived quality of commuter: the perceived

travel time, safety, security, punctuality and comfort are factors influence to mode choice and reason to choose bus of commuter are chosen for studying

- Walking behavior: Change in walking distance, walking attitude and body gesture

- Trip chaining: Change in trip chaining arrangement and multi-purpose trips because

one of reason that the commuter use private vehicle is full fill their daily high trips

Bus user income 2016

<2 mil VND 2-4 mil VND 4-6 mil VND 6-10 mil VND >10mil

CHAPTER IV: HYPOTHESES:

According to the TRAMOC BRT 3-month operation report, the average operating) On

average, BRT riders save 14 percent (or 3.1 minutes in the 9th month BRT survey) of their travel time (including walking time to the station and waiting time at the station) per trip For those who shifted from regular buses to BRT, their travel time savings is much higher: 37–87 percent as the higher frequency and reliability of BRT buses further saves waiting time for riders So that with the advantages of BRT in LOS and operation, the commuter will have the positive change on behavior compare when using BRT than using local bus

Hypothesis 1: When explore people‟s travel behavior, Vu Anh Tuan (2015) claimed that

BRT not effective enough to attract large of number motorcycle and car user until MRT will

be introduced and majority bus user is low income commuter Beside that, the motorcycle take a dominant role in group user that has age from 25 to 60 (Vu Anh Tuan, 2013) According to another evidence, in “Developing the public transport – Lessons learnt from bus system operation practice in Ha Noi” (Vu Hong Truong,2016) found the same opinions that the bus users are mostly student/pupil (37%), the office only take a small part (12%) and the low income group (below 6 mil VND) account for high proportion (81%), the group has income larger than 10mil VND only take 5%

However, Jain et al (2014) found that commuters are willing to shift to public transport if major criteria of services are fulfilled Thanned Satisenam (2016) also claimed that people had positive opinions about BRT as new, better alternative mode of travel

Then with BRT implementation, with the advantages of BRT compare with normal bus will attract more people in other groups bus and change on characteristic of commuter: higher income, the office commuter, group of people have age above 25 and attract more motorcycle commuter than bus (the number of motorcycle shift to BRT larger than bus

Hypothesis 2: Thanned Satisenam (2014) found that travel time and travel cost significantly

affected mode shift to BRT and the reason to use bus (public transport) are cheap and safety although its time is longer than using private vehicle Vu Anh Tuan (2015) also stated that low income commuter, they chose bus for cost saving and the higher (middle and high income) chose motorcycle and car for time saving and comfort respectively So the travel

cost is affected to mode choice of commuter when using bus

Hypothesis 3: Many research proved the relationship with level of service (LOS) with the

commuter‟ satisfaction (Eboli and Mazzulla, 2007) Satoshi Fuji and Hong Tan (2009) also found that the effect of bus service on the intention of using the bus of motorbike users, thus enhancing the "feel" of the quality, can make the intention of using the bus become better However, Fuji (2001) stated that the private vehicle user has the negative opinions about public transport and Vu Hong Truong (2016) found that the perceived quality for bus of commuter is decreasing from 2010 because bus quality is not satisfying when the income and passenger variability and desire for better quality

With the advantages on travel time of BRT, the weight of travel time is higher than travel cost for mode choice of commuter

With higher LOS, The perceived service quality of BRT (travel time, safety, punctuality…) commuter will become better than perceived service of bus commuter

in order to using public transport

11

Hypothesis 4: Statistical data shows an increase in obesity and overweight rates and a

decrease in walking rates throughout the world and in Asian countries (Ebru Cubukcu*,

2013 According report of ADB on walkability and Pedestrian facitilies in Asian country

2011, the walking reducing because of private vehicle domination and polluted environment Another reason for decrease in walking is the domination of motorcycle which often use walkways as parking lots or as way to avoid traffic jams However, improve public transit can lead to improve transit walking and physical activity (Amy L Freeland, Shailendra N Banerjee, Andrew L Dannenberg, Arthur M Wendel,2013)

- H

y

pothesis 5: Vu Anh Tuan (2015) found that people in developing countries have high

trip per day than developed countries, the average trip number is 4 trip/ day then they use motorcycle to fit their demand on trip chaining and saving time

CHAPTER V: METHODOLOGY AND APPOARCHES

5.1 Methodology:

To impact evaluation that is Observe the change in bus commuter behavior (Modal shift, Perception, walking behavior, trip pattern) before (Y|P=0) and after (Y|P=1) when BRT was implemented:

)0

|()1

|(   



However: Simply observing the change will not give the properly causal impact because many factor are likely to influence the commuter over time then to observe the changes and the impact evaluation (Modal shift, Perception, walking behavior, trip pattern) on commuter behavior this study using counterfactual methods compare the treatment group (BRT commuter) and comparison group (local bus commuter)

With higher LOS, BRT encourage the walkability of commuter than bus commuter

in term of walking distance, walking attitude

When using BRT, the commuter will arrange their trip changing to suit the BRT trip and decrease number of trip per day

To evaluate exactly the change on commuter behavior this study aims to use 2 counter factual method “change over time” and “participate and not participate”:

1.DID: Difference in Differences method (before and after evaluation) to observe changes

of BRT commuter (treatment group) “before and after “ BRT implementation including consider the change of bus commuter (comparison group)

2 : change overtime of comparison group

(𝑌|𝑃 = 1) : BRT commuter in examined time

(𝑌|𝑃 = 0) : BRT commuter before implementation (when using bus)

(𝑌 |𝑃 = 1) : bus commuter in examined time

(𝑌 |𝑃 = 0) : bus commuter before implementation

The difficulties of this method is the commuters must recall the information of the moment before BRT implementation (2017) Another difficulty of this method is different characteristic and context of buses and BRT commuter is identified the trend in changing of bus user

𝜟𝟏 = (𝒀𝟏|𝑷 = 𝟏) − (𝒀𝟏|𝑷 = 𝟎) - 𝜟𝟐

𝜟𝟐 = (𝒀𝟎|𝒑 = 𝟏) − (𝒀𝟎|𝑷 = 𝟎)

BRT EFFECT

Data of BRT commuter in 2018 (investigation time)

er

Bus

commut

13

2 PSM (Propensity Matching Method): participate and not participate evaluation

Observe differences of BRT commuter (particitpate on using BRT) with another local bus commute (not participate on BRT) Compare similar groups of BRT and Bus commuter that have similar predicted probability of choosing bus (propensity score) in other words compare BRT and bus commuter which has same common support which determined by propensity score

Estimate the propensity score: P D( i  1|X i) G(0 1X i)

X: variables that effect the predict probability of choosing bus such as: gender, job, income, vehile used in 2016, ages

After that matched sample (commuter) has same propensity score in common support area and fine the difference

BRT commuterBus commuter

The difference BRT commuter (𝑌|𝑃 = 1)

The impact of BRT implementation on commuter will determined by the different of BRT commuter and the bus commuter

The difficulty: different context (characteristics) between BRT route and other bus routes and require collect all variables that effect to mode choice of commuter

5.2 Sampling design

In this study, the treatment group is the BRT commuter and the control group is the bus user However, the difficulty is the identified the trend in changing of bus user and different context between local bus route To solve this problem this study will examine 3 difference local bus routes to comparison Total of the sample number for each bus is 200/ route The choosing bus route is similar to each other in order of characteristic such as: go through high demand areas, bus route in main road of city with cross section > 4 lanes, same fee

5.3 Questionnaire design

Two type of survey is design: one for the BRT commuter and the other for the local bus commuter Both of type has questionnaire in two period of time: 2016 (before BRT implementation) and 2018 (after BRT implementation)

The structure of survey was divided into 2 part:

𝐴𝑇E = (𝑌|𝑃 = 1) − (𝑌 |𝑃 = 1)

15

Part 1: Commuter basic information such as: gender, age, income, vehicle using on

2016, travel time, travel distance, number of time using bus

Part 2: Commuter information about the perception when using bus/BRT, walking behavior and trip chaining The perception data on LOS and walking behavior is collect by

5 point scale method which has 1 point is very bad, 2 point is bad, 3 is normal or neither bad nor good, 4 is good, 5 is very good The perception on LOS are perceived total travel time, safety, punctuality, security, comfort, satisfaction, chose_time (commuter need chose departure time to avoid rush hour), child <11 (commuter allow child to using bus) and perception of commuter when comparing with using motorcycle which affect to mode choice of commuter The 5 point scale method will apply for evaluate the walking behavior

as well In term of walking behavior, this questionnaire will ask the commuter in rush walking (which mean when using BRT/bus they need to rush for get on/ get off bus or catching bus , the high point equal to negative point for commuter) ,walking_costume (when using BRT/bus commuter will chose the costume fit to their traveling and formal), walking_prefer (when using BRT/bus commuter will have more positive toward walking)

The stated choice was conducted on survey between travel time and travel cost to determine which attribute more effect to mode choice of commuter between motorcycle

In trip chaining, commuter will be asked about how many trips per day and which mode of transport they use for their trip and how they combine their trip when using BRT/Bus to identify which mode help commuter arrange their trip in daily life

CHAPTER VI: CASE STUDY

6.1 Context background:

Ha Noi city is the capital of Viet Nam and one of MDCs which heavily dominated by motorcycle and suffered form congestion and pollution (70% pollution form the traffic) The city manage made many solution to restrict motorcycle and private vehicle However, The number of motorcycle increase each year and the number of bus user decreasing

Meanwhile Hanoians were less motivated to use public transportation even they are living

in a polluted environment caused by private vehicles and road users who currently owned a private vehicle are not keen to use public buses They have low intention to switch from their „motor biking‟ lifestyle (Poh yen Ng and Phung Thanh Phuong,2015) The characteristic of commuter are mostly student and low income according to Vu Anh Tuan (2015), Vu Hong Truong (2016) The walkability of Ha Noi commuter is declining and poor infrastructure forces people to abandon walking and cycling and use motorcycles instead (according to ADB report,2011)

In order to improve the transport system and encourage people using public transport, Ha Noi was implemented first BRT 01 route (Kim Ma – Yen Nghia) in 01/2017 however there are still exist many mixed opinion about the effectiveness of BRT project, so that this case study is representative

BRT Project - Line No.1 Yen Nghia - Kim Ma:

0246

Private Vehicles of Ha Noi

( Ha Noi Traffic Police Department, Transportation

17

The BRT project is one of three components of the World Bank's (WB) Urban Transport Development Project (WB), which aims to develop a pilot public transport by BRT to enhance the transport capacity of the public passenger transport system, meeting the needs of the people on a very important transport axis of the city from the Yen Nghia to Kim Ma (BRT bus from Kim Ma to Yen Nghia bus station, 14.7 km long and follow the route of Kim Ma - Giang Van Minh - Giang Vo - Lang Ha - Le Van Luong - To Huu - Le Trong Tan - Quang Trung (Ha Dong) - Ba La - Yen Nghia Bus Station), contributing to improving the quality of public transport services, public transportation, restricting personal vehicles, reduced traffic congestion and gradually improve the quality of the city's environment This is a new type of public passenger transport piloted in Vietnam but has been successfully applied in many countries around the world

BRT Line No.1 Yen Nghia - Kim Ma

- Route in high demand Road (go through old districts to new districts)

- Road cross-section >4 lanes

- Same fee

a Local bus routes for survey

High demand

District

Long Bien, Hoan Kiem, Hai Ba Trung, Dong Da, Thanh Xuan,

19

Bus Route 30 (Mai Dong – My Dinh)

- The Bus commuter data was collected in

4 high demand areas for each bus include 2 terminals, old district area and in area around intersection with main ring roads (RR2, RR3) The survey was conducted in bus stops and bus

in these areas

b BRT route for survey

- The data of BRT commuters were collected in high demand areas for each bus include 2 terminals, old district area and in area around intersection with main ring roads (RR2, RR3) The survey was conducted in bus stops and BRT in these areas

6.3 Data Analysis

6.3.1 BRT commuter data:

- BRT commuter overview

Deviation N Minimum Maximum Mean

Std Deviation

mean service quality and satisfied perception of BRT user above 4 point (good) compare

with around 3 (neither good nor bad) of bus use only the variable chose-time perception is

the same because almost commuter in MDCs tend to avoid the congestion In compare

speed with motorcycle, the commuter evaluates that BRT faster than motorcycle but the bus

is not with mean 3.56 for BRT and 2.34 for bus In Walking behavior aspect show up a

optimistic that BRT encourage the commuter have better walking behavior: further walking distance (mean is 526.65 compare with 413.02 and mean on walking longer is 1.69) and do not rush when walking to bus stop (mean 2.73 compare with 3.41) In trip chaining pattern, the trip number variables are not different between 2016 and 2018 but the combine trip

variables is significant different maybe the commuter consider BRT more convenient than

bus commuter

- Compare the change on BRT commuter from 2016 to 2018:

21

Paired Samples Test on BRT commuter between 2016 and 2018

Paired Samples Statistics

Mean N Std Deviation Std Error Mean Pair 1 Travel time perception2016 3.52 113 683 064

Pair 2 Punctuality perception2016 3.17 113 767 072

Paired Differences t df Sig

(2-tailed) Mean Std

Deviation

Std

Error Mean

95% Confidence Interval of the Difference Lower Upper

Pair 1 Travel time perception2016 - Travel

time perception2018 -.708 .842 .079 -.865 -.551 -8.940 112 .000

Pair 2 Punctuality perception2016 -

Punctuality perception2018 -.991 .977 .092 -1.173 -.809

10.780 112 .000

-Pair 3 Sercurity perception2016 - Security

-Pair 6 Child<11 perception2016 - Child<11

12.510 112 .000

-Pair 7 Chosetime perception2016 -

-Pair 9 Compare_speed_motor2016 -

Compare_speed_motor2018 -1.207 1.084 .101 -1.406 -1.008

11.996 115 .000

Deviation N Minimum Maximum Mean

Std Deviation

The results got from 192 valid samples in Route Bus 01 commuter survey The data

showed that female prefer bus than male, the average age and income of bus commuter are

30 and 5 mil VND respectively The mean data showed up there are a little different

between year 2016 and 2018 so it is not significant (2018 is better) Specially, the walking

distance showed that the commuter‟s acceptable walking distance in 2018 is shorten than in

2016 and numbertrip, combine_trip variables showed up bus commuter has increase trip

number per day in 2018

Compare the change

on Bus 0 1

commuter from 2016

to 2018

Paired Samples Statistics

Mean N Std Deviation Std Error Mean Pair 1 Travel time perception2016 3.78 145 640 053

Pair 2 Punctuality perception2016 3.70 145 670 056

(2-Std

Deviation

Std Error Mean

95% Confidence Interval of the Difference Lower Upper Pair 1 Travel time perception2016 -

Travel time perception2018 -.186 .656 .054 -.294 -.079

3.418 144 .001

-25

The Sample T test for 145 pair samples show up there is different between data

on Bus 01 commuter on 2016 and 2018 on pair 1,3,5,8,11,16,17(time_perception, security, satisfied_ perception and number_trip, combinetrip)

N Minimu

m Maximum Mean

Std Deviation

Pair 2 Punctuality perception2016 -

Punctuality perception2018 -.090 .745 .062 -.212 .033

1.450 144 .149 Pair 3 Sercurity perception2016 -

-Security perception2018 -.166 .646 .054 -.271 -.060

3.087 144 .002 Pair 4 Safety perception2016 - Safety

1.744 144 .083 Pair 5 Comfort perception2016 -

-Comfort perception2018 -.182 .613 .051 -.283 -.081

3.550 142 .001 Pair 6 Child<11 perception2016 -

-Child<11 perception2018 -.076 .678 .056 -.187 .035

1.348 144 .180 Pair 7 Chosetime perception2016 -

-Chosetime perception2018 -.117 .829 .069 -.253 .019

1.703 144 .091 Pair 8 Satisfied perception2016 -

-Satisfied perception2018 -.276 .651 .054 -.383 -.169

5.105 144 .000 Pair 9 Compare_speed_motor2016 -

-Compare_speed_motor2018 -.076 .515 .043 -.160 .009

1.774 144 .078 Pair 10 Compare_security_motor2016 -

-Compare_security_motor2018 -.055 .537 .045 -.143 .033

1.237 144 .218 Pair 11 Compare_satisfied_motor2016 -

-Compare_satisfied_motor2018 -.117 .661 .055 -.226 -.009

2.134 144 .035 Pair 12 Walking_distance 2016 -

2.699 140 .008 Pair17 CombineTrip2016-

-CombineTrip2018 -.353 1.919 .140 -.630 -.076

2.516 186 .013

tor 124 1 5 2.54 0.914 183 1 5 2.64 0.878 Compare_security_

2016 however, the walking distance showed that the commuter‟s acceptable walking

distance in 2018 is shorten than in 2016 as bus 01 Otherwise, the mean of number of trip and combine_trip variables are different compare with bus 01 commuter that showed

unchanged between 2016 and 2018

Compare the change on Bus 30 commuter from 2016 to 2018

Paired Samples Statistics

Mean N Std Deviation Std Error Mean Pair 1 Travel time perception2016 3.77 124 688 062

Pair 2 Punctuality perception2016 3.74 124 673 060

(2-Std

Deviation

Std

Error Mean

95% Confidence Interval of the Difference Lower Upper Pair 1 Travel time perception2016 -

Travel time perception2018 -.177 .776 .070 -.315 -.039

2.545 123 .012 Pair 2 Punctuality perception2016 -

-Punctuality perception2018 -.145 .852 .077 -.297 .006

1.896 123 .060