AN EMPIRICAL ANALYSIS OF BEIJING PUBLIC TRANSPORT

USERS’ SATISFACTION LEVEL

Hongmei Wang and Lingyu Jia

School of Economics and Management, Beijing Jiaotong Universtiy, Haidian district, Beijing, 100044, China

Keywords: Public transport, Users’ satisfaction, Matter element model, Coefficient of variation.

Abstract: Assessing the level of public transport users’ satisfaction is important not only to implication of public

transport priority policy, but also to improvement of public transport service level. In this paper, evaluation

index system of pubic transport passengers’ satisfaction is established according to basic requirement of

passengers. Then, based on matter element theory, evaluation model which uses coefficient of variation

method to calculate the weight of the evaluation indices is developed. Finally, the model is applied to

Beijing based on data surveyed by questionnare. The research indicates that public transport user

satisfaction degree in Beijing is ‘medium’.

1 INTRODUCTION

Along with the too rapid growth of motor vehicles,

congestion in Beijing has long been an increasingly

serious problem which brings negative effect to

urban development and daily life of residents.

Because of the limit in land resources, congestion in

Beijing can not be solved by increasing the area of

road substantially. As a result, transportation

demand management (TDM) emerged as a useful

tool. Improving public transport is one of the main

TDM measures. Preferential development of urban

public transport is an important measure to raise the

utilization rate of transport resources and reduce

traffic congestion. A convenient, fast and

comfortable public transport system can attract more

residents and thus improve the urban transport

structure. Whether residents choose public transport

mainly depends on its performance. The public

transport system should be improved according to

the requirements of the passengers. In that way

public transport can be more attractive to the

residents, so the congestion can be relieved effetely.

Assessing the level of public transport user

satisfaction is not only important to the implication

of public transport priority policy, but also of great

significance to the improvement of the public

transport service level. For the above reasons, this

paper focuses on the evaluation of public transport

user satisfaction.

2 LITERATURE REVIEW

Evaluation of public transport is worldwidely

academic concerned due to practical significance.

Foreign scholars have a preference for

questionnaire-based survey and statistics to analyze

factors affected to public transport user satisfaction.

Based on data of SP survey, Hensher & Stopher

(2003) did a research on 13 factors’ influence on

passengers’ satisfaction; the 13 factors include bus

travel time, seat availability on bus, driver attitude

and general cleanliness on board etc. Tyrinopoulos&

Antonio (2008) and Olio et al (2010) did similar

analyses on influencial factors.

Some scholars focus on evaluating methods of

public transport performance. Yeh et al (2000)

developed a fuzzy multicriteria analysis model to

assess the performance of bus companies; the model

was then applied to evaluate ten bus companies’

perfrmance in Taiwan. Cheng & Wang (2009)

established an evaluation system based on

government, transit operators and passengers, city of

Zhengzhou was selected for the empirical study.

Some scholars assessed public transport

performance by grey theory method. Li & Hu (2006)

presents 23 evaluation indices involve infrastructure,

384

Wang H. and Jia L..

AN EMPIRICAL ANALYSIS OF BEIJING PUBLIC TRANSPORT USERS’ SATISFACTION LEVEL.

DOI: 10.5220/0003586003840388

In Proceedings of the 13th International Conference on Enterprise Information Systems (SSSCM-2011), pages 384-388

ISBN: 978-989-8425-54-6

 2011 SCITEPRESS (Science and Technology Publications, Lda.)

service and benefit; and the city of Qingdao was

studied by means of grey clustering method. Chen &

Zhang (2009), Li & Sun (2010) did similar research

on Lanzhou and Jinan. Shao et al (2009) established

a comprehensive evaluation system based on

analytic hierarchy process (AHP) and grey relational

analysis in evaluation of public transport of

Yinchuan. Based on questionaire surveyed data,

Yang & Chen (2005) analyzed the influence degree

of attributes like road density, average speed,

departure intervals and accident rate on users’

satisfaction.

In the model built by Shao (2005), indices

involved five aspects including public transport

infrastructure, investment of bus companies, public

transit capacity and service quality. After the indices

are weighted by means of analytic hierarchy process

(AHP), evaluation was realized by an improved BP

neural network model.

Overview of literature above indicates that

evaluation criteria established in most of the recent

research mainly used data related to public transport

system instead of passengers’ subjective feeling.

Furthermore, fuzzy multicriteria analysis, grey

theory method and BP neural network model is

widely used to evaluate the service quality of public

transport. However, fuzzy multicriteria analysis and

grey theory method is often criticized for the

definition of membership function and

whitenization

function is arbitrary. BP neural

network model is

only effective when mass typical data is available.

So, in this paper, the authors will promote public

transport users’ satisfaction evaluation criteria from

the perspective of passengers. In addition,

comprehensive evaluation model is established

based on the theory of matter element.

3 EVALUATION MODEL OF

PUBLIC TRANSPORT USERS’

SATISFACTION LEVEL

Matter element analysis is an appropriate tool in

solving complex and incompatible problems.

Multicriteria evaluation model is established based

on matter element theory and it can be applied to

evaluate the public transport user satisfaction.

A set of n indices



CC C"，

are chosen to

evaluate public transport user satisfaction and the

corresponding numerical values of indices

are

XX"，

. The public transport user

satisfaction can be expressed as a matter-element:

()

XCN

XCNR

,, ==

R represents matter-element of classic domain

while

N is the jth grade of satisfaction and

ijiji

⎡

⎤

⎣

⎦

is the value range of jth grade of

satisfaction on the ith index.

is matter-element

of section domain while P is whole grades of

satisfaction and

[

]

Pi Pi Pi

ab=

represents the

value range of

[

]

[]

jnjn

jjj

baC

baCN

222

111

[

]

[]

111

222

nPnPn

PC ab

Cab

Correlation degree is defined as the membership

between the index and the grade. In extenics,

correlation degree can be calculated by the

correlation function bellow:

()

(

)

()

()()

iji

ipi iji

xx xx

ρρ

⎧

⎪

−∈

⎪

⎨

⎪

∉

⎪

−

⎪

⎩

where



()

jijijiji

ijii

abba

xxx

−

−=

jijiji

bax −=

and

(

)

is the correlation degree

between the ith index and the jth grade. The

correlation degree of the evaluation object

matter-element and the jth grade can be calculated

as:

() ()()

xwxkxL

∑

where

(

)

wx is the weight of the ith criterion.

4 EVALUATION SYSTEM OF

MUNICIPAL PUBLIC

TRANSPORT USERS’

SATISFACTION

The evaluation index system is established follow

the principles of systematicity, multi-levels,

AN EMPIRICAL ANALYSIS OF BEIJING PUBLIC TRANSPORT USERS' SATISFACTION LEVEL

385

Table 1: Evaluation indices, values and weights of Beijing public transport users’ satisfaction level.

Satisfaction

level

Criteria Sub-criteria and their Weights Mean value of Indices

Weights of Indices

Efficiency 0.283

Waiting time at offpeak hours 0.156 7.41 0.044

Waiting time at peak hours 0.166 7.07 0.047

Transfer time 0.158 7.37 0.045

Travel time 0.275 3.74 0.078

Traffic information 0.245 4.57 0.070

Convenience

0.263

Transfer times 0.220 5.96 0.058

Walking time from station to destination 0.396

4.58 0.105

Station information by broadcast 0.174 6.23 0.046

Convenience during transferring 0.209 5.79 0.055

Comfort

0.361

Waiting order 0.169 5.12 0.061

Vehicle cleanliness 0.112 5.87 0.040

Vehicle temperature and air condition 0.148 5.15 0.053

Degree of crowding in the vehicle 0.241 3.23 0.086

Running stability of vehicle 0.200 4.77 0.072

Seat comfort 0.130 5.95 0.046

Service quality

0.093

Staff friendliness 0.486 5.85 0.045

Offerin

seats to o

her

erson 0.514

6.28 0.048

scientificity and feasibility. Finally, a three-level

evaluation index system was established. In this

system, municipal public transport user satisfaction

is the evaluation objective; five attributes including

efficiency, convenience, comfort and service are

considered as criteria; 17 sub-criteria are selected in

the third level (see table 1).

5 EMPIRICAL ANALYSIS OF

BEIJING PUBLIC TRANSPORT

USERS’ SATISFACTION

In this part, we will evaluate the satisfactory level of

Beijing public transport users based on the above

questionaire surveyed data.

5.1 The Questionaire Survey

A questionnaire survey was conducted at bus station,

urban rail and subway station, public transit hub,

shopping center, parks and schools etc. in Beijing in

January 2010 to obtain data of indices. The

questionnaire was designed based on the evaluation

index system. Indices were translated into questions

according to the actual situation to get information

from the passengers. We sent out 700 questionnaires

and 527 of them were collected. Of the 527

correspondents, 252 persons are female while 257 of

them are male; 182 (34.5%) correspondents’ families

have at least one private car; as to age structure, 443

(84%) of them are 18-45 years old, while senior

persons (elder than 60 years old) only occupied

0.9%.

Passenger’s judgment about public transport is

described as five grades, ‘very poor’, ‘poor’, ‘fair’,

‘good’ and ‘very good’ and the corresponding

numerical values are 1, 3, 5, 7 and 9 when recording

the data. The statistical data is presented in Table 1

(see column ‘mean value of the indices’).

5.2 Weights of Indices

Coefficient of variation method can be used to

calculate the weighs of indices objectively. It is

adopted in this paper to avoid the defect of

subjective methods such as AHP and Delphi method.

The weight can be calculated as follows:

(1)

ICEIS 2011 - 13th International Conference on Enterprise Information Systems

386

Table 2: Relational degrees in Beijing public transport

evaluation system.

Relational

degree

Ve r y

good

Good Medium Poor

Very

poor

Grade

Efficiency -0.4276 0.021 -0.159 -0.313 -0.571 Good

Convenience -0.435 -0.108 0.128 -0.294 -0.509 Medium

Comfort -0.489 -0.247 0.163 -0.105 -0.448 Medium

Service

quality

-0.396 0.050 -0.011 -0.414 -0.582 Good

Synthetically

relational

degree

-0.449 -0.119 0.046 -0.242 -0.511 Medium



∑

(2)

Where



is the mean value of jth index,



the standard deviation of the jth index,

is the

coefficient of variation and

is the weight of the

jth index.

Weights calculated by means of coefficient of

variation method are presented in Table 1 (see

column ‘weights of indices’).

5.3 Evaluation of Public Transport

users Satisfaction Level

Mean value of the satisfaction degree obtained by

the survey is taken as the value of corresponding

index. Public transport user satisfaction degree can

be described as five grades, ‘very good’ (the value

range is 8-9), ‘good’ (6-8), ‘medium’ (4-6), ‘poor’

(2-4) and ‘very poor’ (1-2). According to the model

built in part 2, matter-element of evaluation object,

matter-element of classic domain and matter-element

of section domain can be defined and relational

degree between the indices and five grades can be

calculated. Synthetically relational degree between

public transport user satisfaction and five grades is

obtained by weight sum. According to the definition

of relational degree, the grade of public transport

user satisfaction is given by



)(

,...2,1

max x

. The

relational degrees are presented in Table 2.

Data in Table 2 indicates that public transport

passenger satisfaction degree in Beijing is ‘medium’.

Furthermore, synthetically relational degree about

‘good’ is larger than it of ‘poor’. According to the

definition of relational degree in extenics, it is more

likely to translate into grade of ‘good’ from

‘medium’.

6 CONCLUSIONS

By improving public transport service quality to

satisfy passengers, more residents will be appealed

to use public transport instead, so that the congestion

could be relieved. In order to evaluate the public

transport user satisfaction, this paper proposed an

evaluation model based on matter-element analysis

theory. Based on the data obtained by questionnaire

survey, the public transport user satisfaction is

assessed from the perspective of passengers.

Coefficient of variation method is adopted to

calculate the weights of indices to overcome the

shortcoming of subjective methods used in previous

research. The research indicates that public transport

user satisfaction degree in Beijing is ‘medium’ and

close to ‘good’. Poor user satisfaction about

convenience and comfort affect the overall

satisfaction of public transport service. While

striving to construct public transport infrastructure,

the government should pay more attention to

improve the service quality of public transport to

make public transport more attractive.

*Supported by “the Fundamental Research Funds for the

Central Universities (Appraisal of TDM Measures in

Occurring Urban Congestion)” and “National Nature

Science Fund of China (Research of residents’ Selective

Mechanism in Public Transport)”.

REFERENCES

Chen Han, Zhang Xiaoyuan. Comprehensive evaluation of

the operation level of urban public transportation

based on the travel requirement of residents [J].

Technology & Economy in Areas of Communications,

2009, 6: 8-11.

Cheng Xi, Wang Wei, Ren Gang et al. A study on

comprehensive evaluation of metropolitan public

transit with fuzzy evaluation and analytic hierarchy

process [J].

Urban Public Transport, 2009, 2: 25-29.

Hensher D., Stopher P., Bullock P., Service

quality-developing a service quality index in the

provision of commercial bus contracts [J].

Transportation Research (Part A), 2003, 37: 499-517.

Li Yinghong, Sun Huijuan. Evaluation and research of

BRT service level[J]. Journal of Chongqing Jiaotong

University (Natural Science)

, 2010, 29(2): 285-289,

AN EMPIRICAL ANALYSIS OF BEIJING PUBLIC TRANSPORT USERS' SATISFACTION LEVEL

387

290.

Li Yuhua, Hu Yunquan. Gray clustering method applied to

evaluation of urban public transport development level

[J].

Mathematics in Practice and Theory, 2006, 32(6):

125-132.

Olio L, Ibeas A, Cecín P. Modeling user perception of bus

transit quality [J]. Transport Policy, 2010, 17:

388-397.

Shao Fei, Deng Wei, Yi Fujun et al. Comprehensive

evaluation method of metropolitan public

transportation system based on analysis hierarchical

process and grey theory[J]. Journal of PLA University

of Science and Technology (Natural Science Edition)

2009, 10(6): 536-541.

Shao Zufeng. Evaluation model of urban public traffic

service quality based on neutral network [J]. Municipal

Administration & Technology,

2005, 4: 178-180.

TyrinopoulosY, Antoniou C. Public transit user satisfaction:

Variability and policy implications [J]. Transport

Policy,

2008, 15: 260-272.

Yang Jun, Chen Rongqiu, Guo Congmin. Assessing

service quality of city public transport via grey

relational theory [J]. Industrial Engineering and

Management,

2005, 10(4): 89-92.

Yeh C, Deng H, Chang Y. Fuzzy multicriteria analysis for

performance evaluation of bus companies [J].

European Journal of Operational Research, 2000,

3(126): 459-473.

ICEIS 2011 - 13th International Conference on Enterprise Information Systems

388