Evaluation of Customs Service Quality by Using Fuzzy SERVQUAL and Fuzzy MCDM

Abstract

Service quality has been revealed as a key factor in search for sustainable competitive advantage, differentiation and excellence in the service sector. Customers' evaluations of the service quality are critical to service firms that aim to improve their marketing strategies so Accurate measurement of service quality is a major concern to management. But, whereas measuring criteria of services quality and satisfaction are fuzzy and ambiguous but available methods measuring them generally is classic kind. So Appling fuzzy methods according to other methods are closet to human thinking. Therefore¸purposeTherefore¸Therefore¸purpose of this paper is service quality evaluation of service organizations by using Fuzzy MCDM approach. The paper points out a new insight of fuzzy multi-criteria decision making techniques to evaluate service quality that involved three techniques (Entropy , Fuzzy Servqual , Topsis) and finally we provides an example from Iran's customs to illustrate how the procedure is used.

Full text

American Journal of Scientific Research
ISSN 1450-223X Issue 35 (2011), pp.89-103
© EuroJournals Publishing, Inc. 2011
http://www.eurojournals.com/ajsr.htm
Evaluation of Customs Service Quality by Using Fuzzy
SERVQUAL and Fuzzy MCDM
Mohammad Ali Abdolvand
Dep.of Business Management, Science &Research Branch
Islamic Azad University (I.A.U), Tehran, IRAN
E-mail: Abdolvand-MA@yahoo.com
Tel: 009809121119125
Mohammad Javad Taghipouryan
Corresponding Author, Dep. of Business Management
Science &Research Branch, Islamic Azad University (I.A.U) ,Tehran, IRAN
E-mail: jpouryan@yahoo.com
Tel: 009809112562789
Abstract
Service quality has been revealed as a key factor in search for sustainable
competitive advantage, differentiation and excellence in the service sector. Customers’
evaluations of the service quality are critical to service firms that aim to improve their
marketing strategies so Accurate measurement of service quality is a major concern to
management. But, whereas measuring criteria of services quality and satisfaction are fuzzy
and ambiguous but available methods measuring them generally is classic kind. So Appling
fuzzy methods according to other methods are closet to human thinking.
Therefore ¸ purpose of this paper is service quality evaluation of service
organizations by using Fuzzy MCDM approach. The paper points out a new insight of
fuzzy multi-criteria decision making techniques to evaluate service quality that involved
three techniques (Entropy , Fuzzy Servqual , Topsis) and finally we provides an example
from Iran’s customs to illustrate how the procedure is used.
Keywords: Service quality -Fuzzy SERVQUAL- Entropy - Topsis.
1. Introduction
As most developed economies are now service, rather than product-oriented, industries, service quality
holds a prominent position in the marketing/management literature (Chau & Kao, 2009).
Service quality has been revealed as a key factor in search for sustainable competitive
advantage, differentiation and excellence in the service sector (Jabnoun and Al Rasasi, 2005; Jun et al.,
1998). Also, it has been recognized as being highly important for satisfying and retaining customers
(Spreng et al., 1996; Reichheld and Sasser, 1990).
Customers’ evaluations of the service quality are critical to service firms that aim to improve
their marketing strategies. Firms that provide superior service quality also have a more satisfied
customer base. Customer satisfaction is viewed as influencing repurchase intentions and behavior,
which, in turn, leads to an organization’s future revenue and profits (Taghipouryan &et al, 2010).

Evaluation of Customs Service Quality by Using Fuzzy SERVQUAL and Fuzzy MCDM 90
Service quality concerns the benefits on the customer side of the company-customer exchange.
Carefully integrating issues of quality with productivity, company improvement programs are aimed
generally at improving the long-term profitability of the firm (Lovelock and Wirtz, 2004)
In dealing with a decision process, the decision-maker is often faced with doubts, problems and
uncertainties. In other words Natural language to express perception or judgment is always subjective,
uncertain or vague (Sanayei, 2009).
It is along time that behavioral science scientists and quality theorist have understand, there is
ambiguity in a lot of humans judgment. In 1965, an Iranian professor in Colombia University (Zadeh)
introduced Fuzzy set in humans systems (systems with human interaction such as services systems)
and decision procedure, as a tool against ambiguity and imprecise (E.Abbott, 1996).
Multi-criteria decision-making (MCDM) methods that involve multiple, and usually conflicting
criteria allow decision makers to deal with complex evaluation problems to achieve a certain goal
(Percin, 2010) .Fuzzy set theory quantifies human subjective knowledge, and the key elements in
human thinking are not numbers, but linguistic terms or fuzzy set labels (Zadeh, 1965); therefore, there
is a need to develop a MCDM method, based on fuzzy set theory (Lai & et al, 2010). So, Zadeh and
Bellman [23] developed fuzzy multi-criteria decision-making (FMCDM) methodology to resolve the
lack of precision in assigning importance weights of criteria and the ratings of alternatives regarding
evaluation criteria(Sanayei, 2009).
However, whereas measuring criteria of services quality and satisfaction are fuzzy and
ambiguous but available methods measuring them generally is classic kind (Liou & Chen, 2006) ¸
Appling fuzzy method according to other methods are closet to human thinking.
Therefore ¸ purpose of this paper is service quality evaluation of Iran’s customs by using Fuzzy
MCDM approach. So ¸First, we applied Entropy method for calculating the criteria weights .Then, for
evaluation of Service Quality we used fuzzy numbers on the basis of five dimensions of service quality
in SERVQUAL model which we call Fuzzy SERVQUAL that has five steps .And finally, we conduct
Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) to achieve the final ranking
results. This paper is organized as follows. Section 2 summaries the concepts of SERVQAUL, Fuzzy
logic, Entropy and Topsis. Section 3 describes methodology that included five steps of Fuzzy Servqual,
samples and reliability. Section 4 provides an example to illustrate how the procedure is used. The final
section discusses the results.
2. Lecture Review
2.1. SERVQUAL
Measuring and evaluating service are not considered as add-ons to the overall delivery of service
quality and there is already a well-established wealth of extant literature on measuring service quality,
particularly in relation to customer expectations. The meaning of service quality may vary in different
contexts (for a review, see Ghobadian et al., 1994), but we refer to the view generally understood in
marketing management (Chau & Kao, 2009).
SERVQUAL is one of the most widely used scales in practice that measures customers’
perceptions of SERVQUAL, and it has been shown to be applicable for a wide range of service
industries(Rosenbaum & Wong,2009).
The pioneer study of Parasuraman et al. (1985) has been a major driving force in developing an
increased understanding of and knowledge about service quality (Gerrard and Cunningham, 2001).
They defined service quality as the gap between customers’ expectation of service and their perception
of the service experience. The various gaps visualized in the model are: (Kumar & et al, 2010)
(1) Gap 1: Difference between consumers’ expectation and management’s perceptions of those
expectations, i.e. not knowing what consumers expect.

91 Mohammad Ali Abdolv and Mohammad Javad Taghipouryan
(2) Gap 2: Difference between management’s perceptions of consumers’ expectations and
service quality specifications, i.e. improper service-quality standards.
(3) Gap 3: Difference between service quality specifications and service actually delivered i.e.
the service performance gap.
(4) Gap 4: Difference between service delivery and the communications to consumers about
service delivery, i.e. whether promises match delivery.
(5) Gap 5: Difference between consumers’ expectation and perceived service.
The above four gaps (1 - 4) concern the causes of poor service quality in the way the
organization is managed. The user quality however is measured by gap-5 as the differences between
the expectations and perceptions of customers. Gap-5 depends on the size and direction of the four
disconfirmations associated with the delivery of service quality on the marketer’s side. If gaps 1 to 4
are reduced, then service quality can be improved (Kasper et al., 2006).
Zeithaml and Parasuraman (2004, p. 1) defined SERVQUAL as “the degree and direction of
discrepancy between customers’ service perceptions and expectations.” Furthermore, their research
shows that customers assess SERVQUAL along five perceptional dimensions – assurance, empathy,
reliability, responsiveness, and tangibles – that can be evaluated by the SERVQUAL scale
(Parasuraman et al., 1988). These dimensions are defined as follows:
 Tangibles: physical facilities, equipment and appearance of personnel.
 Reliability: ability to perform the service accurately and dependably.
 Responsiveness: willingness to help customers and provide prompt service.
 Empathy: caring and individualized attention provided to customers.
 Assurance: employees’ knowledge, courtesy and ability to convey trust and confidence.
For each service dimension and the total service, a service quality (Q) judgment can be
computed as perception (P) less expectation (E), i.e. Q = P-E (where Q is represented by gap-5). These
components can be calculated by the same five dimensions – reliability, assurance, tangibles, empathy,
and responsiveness (Chau & Kao, 2009) .Zeithaml et al. (1990) suggest 22 questions (items) that relate
specifically to these five dimensions, on which we base the design of our study.
Numerous studies and investigations on service quality have been carrying out across various
industries after the pioneering work by Parasuraman et al. (1985) such as medical services (Brown and
Swartz, 1989), car retailing (Carman, 1990), travel and tourism (Fick and Ritchie, 1991), hospitality
(Saleh and Ryan, 1992; Johns, 1993), higher education (Ford et al., 1993), accounting firms (Freeman
and Dart, 1993), construction professionals (Hoxley ,2000), public services (Brysland and Curry, 2001;
Agus et al., 2007;Taghipouryan&et al ,2010), and mobile communications (Kung et al., 2009).
2.2. Fuzzy Logic
The logical tools that people can rely on are generally considered the outcome of a bivalent logic
(yes/no, true/false), but the problems posed by real-life situations and human thought processes and
approaches to problem-solving are by no means bivalent (Sanayei, 2009).
In classical set theory, an object is either a member of a set or excluded from it. Thus, in
conventional dual logic, a statement can only be either true or false. In reality, however, human
cognition, perception and judgment involve approximate and vague reasoning, and cannot be modeled
adequately by classical set theory. Fuzzy sets were introduced by Zadeh (1965), as a method of
handling vagueness, or uncertainty, particularly linguistic variables. Fuzzy sets consider the grey area
of data, rather than considering membership of a set to be simply true or false. In other words, fuzzy
sets allow partial membership of a set (Jamali. and Sayyadi, 2009).
For example, in fuzzy logic, a statement such as “Sam is old” can be 60 per cent true and 40 per
cent false; i.e. Sam has a 0.60 degree of membership in a set whose elements are considered to be old
and a 0.40 degree of membership in a set of young people. A person can be a member of both sets at
the same time. Although the sum of the degree of membership is 100 percent in our example, fuzzy
logic does not require it (Beheshti& lollar, 2008).

Evaluation of Customs Service Quality by Using Fuzzy SERVQUAL and Fuzzy MCDM 92
According to the definition of Dubois and Prades [1978], the fuzzy number
A
~
is a fuzzy set,
and its membership function is ]1,0[:)(
~RXu A . It is common to use triangular fuzzy numbers
(TFNs) ),,()( 321
~aaaXu Aor ),,()( 321
~UMLXu A

as shown in Eq. (1).








otherwise
axaaaxa
axaaaax
x
A
0
)/()(
)/()(
)( 32233
21121

(1)
Where a1 and a3 stand for the lower and upper bounds of the fuzzy number
A
~
, respectively; and
a2 stands for the modal value. The operational laws of two TFNs ),,(
~
321 aaaA  and ),,(
~
321 bbbB 
are as follows:
Addition of two fuzzy numbers
),,( 321 aaa ),,( 321 bbb = )3,,( 32211 bababa


 (2)
Subtraction of two fuzzy numbers

),,( 321 aaa ),,( 321 bbb =)3,,( 32211 bababa


 (3)
Multiplication of two fuzzy numbers

),,( 321 aaa ),,( 321 bbb =)3,,( 32211 bababa


 (4)
Division of two fuzzy numbers
),,( 321 aaa ),,( 321 bbb =)3/,/,/( 32211 bababa (5)
2.3. Entropy
For solving MCDM problems, it is generally necessary to know the relative importance of each
criterion. It is usually given as a set of weights, which are normalized, the importance coefficients in
the MCDM methods refer to intrinsic “weight”. The entropy method is the method used for assessing
the weight in a given problem because, with this method, the decision matrix for a set of candidate
materials contains a certain amount of information. In other words, the entropy method works based on
a predefined decision matrix (Shanian & Savadogo, 2006).
The entropy idea is particularly useful for investigating contrasts between sets of data .This
method has it roots in information theory and was introduced in 1948 to provide a quantitative measure
of the “uncertainty” represented by a discrete probability distribution (Soo, 2004).
Entropy analysis is based on three measures: entropy (Ej), degree of divergence (dj), and degree
of influence or weight of importance (wj) that this method consists of the following procedure:
(Shanian & Savadogo, 2006; Kou & Xiong, 2006; Soo, 2004)
 Step 1 : Normalizing the decision matrix :
JjIi
r
r
Pm
iij
ij
ij ,....,2,1,......,2,1
1
  (6)
 Step2: Calculating the entropy with data for each criterion, the entropy of the set of normalized
outcomes of the jth criterion is given by:
JjIippkE m
iijijj ,....,2,1,......,2,1)]ln([
1  (7)
 Step 3: weights of criteria :
j
d
d
w
Ed
n
jj
j
j
jj


1
1
(8)

93 Mohammad Ali Abdolv and Mohammad Javad Taghipouryan
2.4. Topsis
TOPSIS is based on a simple and intuitive concept, it enables consistent and systematic criteria, which
is based on choosing the best alternative having the shortest distance from the ideal solution and the
farthest distance from the negative ideal solution. The ideal solution is regarded as the maximal
benefits solution. It consists of taking the best value of alternative and the negative ideal solution is
treated as the minimal benefits solution, it is composed of all worst value of alternatives.
(Langkumaran and Kumanan, 2009)
The TOPSIS procedure consists of the following steps (Jamali and Tooranloo, 2009;
Chamodrakas & et al, 2009):
Step 1: Construct the normalized decision matrix. The various attribute dimensions are
converted in this step into non-dimensional attributes, in order to allow comparisons across the
attributes. An element of the normalized decision matrix is calculated as follows:
m
i
x
x
rm
iij
ij
ij ,......,2,1
1
2  (9)
Step 2: Construct the weighted normalized decision matrix. The Decision Maker assigns
weights to each attribute. The weighted normalized decision matrix is constructed by multiplying each
element rij with its associated weight wj:
jiijijij
jiij
rwV
VV
,
,
)(


(10)
Step 3: Determine the ideal and the negative ideal solutions. The ideal solution A+ and the
negative ideal solution A+ are defined as:


miJjVjVA
miJjVJjVA
ijij
ijij
,....,2,1),/(min),/(min
,....,2,1),/(min),/(max







(11-12)
Step 4: Measure the separation of alternatives from the ‘‘ideal” solutions. The separation of
each alternative from the ideal and the negative ideal solution is given by:
miVVijS
miVVijS
n
jji
n
jji
,......,2,1)(
,......,2,1)(
1
2
1
2








(13-14)
Step 5: Calculate the ‘‘relative closeness” of each alternative to the ideal solution and the final
ranking. The ‘‘relative closeness” of each alternative to the ideal solution is calculated according to Eq.
(4). It is clear that the following relations hold:
mi
SS
S
C
ii
i
i,......,2,1




(15)
Step 6: Rank the preference order. A set of alternatives can be preference ranked according to
the descending order of 
i
C.
3. Research Methodology
Questionnaire that applied in this research is extracted from criteria of service quality (SERVQUAL).
In this research, questionnaire have been frame in 3 section: first section related to properties of
population , second section is about importance of each 5 dimension by view point of client and in

Evaluation of Customs Service Quality by Using Fuzzy SERVQUAL and Fuzzy MCDM 94
third section there are 22 pairs of items. Half were aimed at measuring service clients' expectations and
the remaining half measured perceptions.
The aim of this section is presentation of a way for evaluation of Service Quality by using
fuzzy numbers on the basis of five dimensions of service quality in SERVQUAL model which we call
Fuzzy SERVQUAL.
Fuzzy SERVQUAL has 5steps which are consisting of:
Step 1- Determination of fuzzy numbers for each of the linguistic variables:
In this study, “strongly disagree” and “strongly agree” five spectrums are used that have been
shown as the following:
Strongly Disagree (SD), (D) Disagree, Middle (M), Agree (A), Strongly Agree (SA)
For gaining each of the linguistic variables’ fuzzy numbers, experts’ opinions were used, so
each expert were asked to determine linguistic variables’ spectrum from 0 to 100.
The sample of these opinions is shown in table (1). Whereas evaluation of custom service
quality is in the clients’ view, we chose experts among clients with at least BA degree and upper and
with the intercourse duration of more than 7years with custom (integration of science and experience)
and we could get this determination of spectrum from 30 persons.
Table 1: Scale of linguistic variables by experts
Scale of linguistic variables(0-100)
SD D M A SA
0 - 20 20 - 40 40 - 60 60 - 80 80 - 100 expert 1
0 - 10 10 - 30 30 - 50 50 - 70 70 - 100 expert 2
. . . . . .
. . . . . .
. . . . .
0 - 20 20 - 30 30 - 40 40 - 60 60 - 100 expert 30
After achieving experts’ opinion by evaluation of these 30 experts in linguistic variables scale,
we determine triangular fuzzy numbers (TFN) of each linguistic variables.
According to the above mentioned, now TFN of each linguistic variables are consist of:
o “Strongly Disagree” linguistic variable (SD) :
So, TFN for SD linguistic variable with membership function is as the following:
Table 2: TFN for SD linguistic variable by using expert’s opinions
L M=(L+U)/2 U
expert 1 0 10 20
expert 2 0 5 10
expert 3 0 10 10
expert 4 0 5 10
expert 5 0 5 10
.
.
.
expert 30
0 10 20
TFN(SD) 0 8.63 20
min average max

95 Mohammad Ali Abdolv and Mohammad Javad Taghipouryan
Figure 1: Triangular membership function of fuzzy number for “Strongly Disagree”
As it was mentioned, we could obtain TFN for SD linguistic variables by experts’ opinion, and
other linguistic variables’ fuzzy numbers are obtained in this way. These numbers with their
membership function are in the following:
SD = (0, 8.63, 20) D = (10, 24.09, 40) M = (20, 38.63, 60) A = (30, 58.63, 80) SA = (60, 85.45,
100)
Figure 2: Triangular membership function of fuzzy number
Step 2 – opinion conversion of each experts who answer the questionnaire according to
the obtained TFN in previous step:
As it was mentioned, data collection tools is a questionnaire based on service quality scale
(SERVQUAL) that operates on the basis of 22 pairs of items relevant to 5dimesion and a comparison
clients expectation and perceptive performance. Since, SERVQUAL is both weighted and non-
weighted, of course it should be mentioned that weighted SERVQUAL was used in this research, the
difference between weighted and non-weighted SERVQUAL in fuzzy numbers is shown in table(2).
Table 2: The difference between weighted and non-weighted forms in fuzzy SERVQUAL in data collection
Level First item Non-weighted fuzzy SERVQUAL weighted fuzzy SERVQUAL
Answer TFN Weight (%) TFN
Expectation A good custom has
modern equipments. Strongly agree (60, 85.45, 100) 90 0.9(60, 85.45,
100)
=(36,76.9,60)
Performance This custom has modern
equipments. Agree (30, 58.63, 80)
90 0.9(30, 58.63,
80)=
(27,52.76,72)

Evaluation of Customs Service Quality by Using Fuzzy SERVQUAL and Fuzzy MCDM 96
In this way, for the given answer to each of the 22 pairs of items of questionnaire, we use
proper TFN which was obtained in the previous step.
Step 3- obtaining fuzzy numbers of each 5dimentions of service quality by using fuzzy
average:
Whereas the present research has 5dimension by using service quality scale, so after opinion
conversation of each expert who answer to fuzzy numbers in each of the questions, this task should be
done in 5dimension level and for this reason fuzzy average is used. In this section both weighted and
non-weighted averages have been shown.
Table 3: Five dimension scale of service quality with correspondence of its items
Dimensions Tangibility Reliability Responsiveness Assurance Empathy
Items correspondence 1 – 4 5 - 9 10 – 13 14 - 17 18 - 22
Number of items 4 5 4 4 5
In the lower part, fuzzy average for dimension of tangibility in two levels of expectation and
performance and with two forms of weighted and non-weighted. Other dimensions are obtained in this
way.





4/)(
4/)(
141312111
141312111
PPPPP
EEEEE
formWieghtNon
T
T (16)
In which:
E1T = expectations of first responder than dimension of tangibility
P1T = perceptible performance in the view of first responder than dimension of tangibility.









)4/()(
)4/()[(
14141313121211111
14141313121211111
wPwPwPwPwP
wEwEwEwEwE
formWieght
T
T (17)
w = Importance coefficient of each items in the view of responders
Step 4 – Defuzzification:
Obtained results in the previous steps for 5dimensions of CSQ have been like TFN that for
analysis and test of hypothesis and comparison between dimensions and decision making should be
changed from triangular number to the crisp number which is called defuzzification.
There are several available methods serve this purpose. Mean-of-Maximum, Center-of-Area,
and a-cut Method is the most common approaches. This study utilizes the Center-of-Area method due
to its simplicity and does not require analyst’s personal judgment (Tsuar &et al, 2002).
The defuzzified value of fuzzy number can be obtained from Eq. (18).
11111
321
3/)]()[(
),,(
LLMLUNFA
UMLA

 (18)
Now in the following part, defuzzification by the way of center of area is for dimensions of
tangibility in two levels of expectations(NFE) and performance(NFP) and other dimensions are also
obtained in this way.
TTTTTT
TTTT
TTTTTT
TTTT
LLMLUNFP
UMLP
LLMLUNFE
UMLE
111111
1111
111111
1111
3/)]()[(
),,(
3/)]()[(
),,(




(19)
Step 5 – Measurement of custom service quality (CSQ):
Whereas in the previous step, we converted fuzzy numbers to the crisp numbers in two levels of
performance and expectation by center of area way. This step for measurement of clients satisfaction,

97 Mohammad Ali Abdolv and Mohammad Javad Taghipouryan
dissatisfaction and delight of CSQ on basis of SERVQUAL model, is obtained by the following
formula:
delightclientsNFSQ
onsaticfacticlientsNFSQ
ctiondissaticfaclientsNFSQ
NFENFPNFSQ
0
0
0




(20)
For determining reliability of this questionnaire from in this research Cronbach's

has been
used. Values of final for each of 5 dimensions of service quality with similar questions are the table 4.
According Sekaran‘s opinion, Cronbach’s coefficient less than 0.6 is weak, 0.7 is acceptable and more
than 0.8 is very good(Amirshahi &Mazhari,2008 ).Therefore the, result of this research for 2
dimensions are acceptable and for 3 dimensions are good and whole questionnaire from have good
reliability.
Table 4: Custom service quality scores: Cronbach’s alpha
Dimensions Tangibility Reliability Responsiveness Assurance Empathy Service quality
Items correspondence 1 – 4 5 - 9 10 – 13 14 - 17 18 - 22 1-22
Number of items 4 5 4 4 5 22
Cronbach’s Alpha 0.73 0.85 0.86 0.83 0.78 0.90
There are eight customs around Caspian Sea in Iran that we have chosen three customs of
Mazandaran province ( Nowshar, Sari, and Amir Abad customs) and 210 questionnaires distributed
between clients ( exporters, importers, commission agents and……) .
Beby (1998) believes 50 percent of answers are enough for analyzing data and result of reports.
Rate of answers for good 60 percent and for very good is 70 percent (Amirshahi &Mazhari, 2008).
In this research, from 210 questionnaires that had been distributed, 11of them (5.25%) hadn’t
been returned , 5 of them(3.28%) weren’t completed and 194 of them were completed that were ready
for analyzing a rate equal with 92.28% that is a good rate.
Table 5: Demographic characteristics of respondents (total number = 194, P= percentages, F= Frequency)
Sex Age Occupation
Intercourse duration with
custom
P F P F P F P F
Male 96.36 187 Below30 49.48 96 Exporter 20.12 39 Below 2 22.16 43
Female 3.61 7 31- 40 35.56 69 Importer 29.89 58 2 – 4 31.96 62
41 – 50 11.34 22 Commission agent 40.72 79 4-7 21.14 41
51 and above 3.61 7 Others 9.27 18 7 and above 24.73 48
4. Data Analysis and Results
4.1. The Weights of Five Dimensions and Twenty Two Criteria
Table 6 shows the relative weights of the five dimensions of service quality, which are obtained by
applying Entropy. The weights for each of the dimensions are:
Tangibility (w = 0.263) - Reliability (w =0.205) - Responsiveness (w = 0.281) - Assurance (w =
0.258) - Empathy (w = 0.211).
The participants in this study have highest weights (or priority) for two item of questionnaire
(w=0.257) "2: The physical facilities at the custom are visual attractiveness" And "12: Staffs at the
custom are always to help you and have lowest weights for two items (w = 0.196) “5: When the custom
promises to do something by a certain time, it did it" and "21: The custom has your best interests at
heart".

Evaluation of Customs Service Quality by Using Fuzzy SERVQUAL and Fuzzy MCDM 98
4.2. Fuzzy Servqual for Service Quality Measure
After obtaining the criteria weights from Entropy (table. 6), by using fuzzy numbers on the basis of
five dimensions of service quality in SERVQUAL model which we call Fuzzy SERVQUAL evaluated
Service Quality of customs. Table 7 lists the Fuzzy SERVQUAL measure for the three customs. Then,
we defuzzified the Fuzzy numbers into crisp numbers so as to conduct TOPSIS ranking procedure. We
used Center-of- Area method to defuzzify the Fuzzy numbers, which are shown in Table 8.
Table 6: Weights of the five dimensions and twenty two criteria.
Five dimensions and twenty two criteria Entropy method
E
j
d
j
w
j
Tangibility (w = 0.263)
The custom has modern – looking equipment. 0.283 0.717
0.251
The physical facilities at the custom are visual attractiveness. 0.265 0.735 0.257
The staffs at the custom appear neat. 0.312 0.688 0.241
Material associations with the service are visually appealing at the custom. 0.289 0.711 0.249
Reliability (w =0.205)
When the custom promises to do something by a certain time, it did it. 0.321 0.679
0.196
When you have problems, the custom shows a genuine interest in solving them. 0.318 0.682 0.197
The custom performs the service right the first time. 0.300 0.700 0.202
The custom provides its services at the time it promise to do so. 0.315 0.685 0.198
The custom insists on error-free service. 0.291 0.709 0.205
Responsiveness (w = 0.281)
Staffs at the custom tell you exactly when services will be performed. 0.315 0.685
0.245
Staffs at the custom give you prompt service. 0.321 0.679 0.243
Staffs at the custom are always to help you. 0.284 0.716 0.257
Staffs at the custom are never too busy to respond to your request. 0.295 0.705 0.253
Assurance ( w = 0.258)
The behavior of staff instills confidence in you. 0.312 0.688
0.247
You feel safe in the delivery of service. 0.294 0.706 0.253
Staffs of the custom are consistently courteous with you. 0.319 0.681 0.244
Staffs of the custom have the knowledge to answer your question. 0.317 0.683 0.245
Empathy ( w = 0.211)
The custom gives you individual attention. 0.293 0.707
0.201
The custom has opening hours convenient to all its clients. 0.300 0.700 0.199
The custom has staffs who give you personal attention. 0.309 0.691 0.197
The custom has your best interests at heart. 0.311 0.689 0.196
The staffs of the custom understand your specific needs. 0.286 0.714 0.203
Table 7: Service quality measures of customs by Fuzzy Servqual
Twenty two criteria custom A custom B custom C
T1 (-65.25 ,-48.72, -31.25) (-54.32, -36.52, -22.16) (-46.32,-29.14, -12.54)
T2 (-8.12,0.00,4.51) (-18.25,-4.39,2.53) (-20.18,-12.45,-4.29)
T3 ( -3.25,1.27,7.38) (0.00 , 0.00 , 4.35) (-1.85, 3.89, 8.92)
T4 ( -18.34,-10.42,-2.16) ( -23.15,-15.86,-8.35) ( -31.24.-24.65,-18.45)
Tangibility (-23.74,-14.46,-5.38) (-23.93,-14.19,-5.9) (-24.89,-15.58,-6.59)
Rel 5 (-62.35,-42.58,-37.21) (-58.27,-41.58,-29.54) (-61.24,-49.28,-32.58)
Rel 6 (0.00,12.32,23.24) (2.45,16.28,28.35) (0.00,0.00,12.34)
Rel 7 (-49.25,-40.29,-28.65) (-51.27,-.39.28,-27.68) (-54.36,-47.25,-37.65)
Rel 8 (-.64.28,-.50.29,-37.68) (-49.28,-38.64,-29.65) (-64.38,-52.28,-43.19)
Rel 9 (-24.56,-13.58, -2.10) (-8.59,0.00,0.00) (-18.33,-9.24,0.12)
Reliability (-40.08,-26.88,-16.48) (-32.99,-20.64,-11.70) (-39.66,-31.61,-20.19)
Res 10 (-67.38,-48.28,-33.29) (-.71.28,-59.28,-43.19) (-69.28,-54.32,-48.37)
Res 11 (-24.91,-13.27,-4.61) (-34.61,-21.34,-14.31) (-35.16,-21.49,-16.25)
Res 12 (-9.84,-1.03,6.24) (-10.24,-2.09,7.31) (-4.31,5.12,12.49)
Res 13 (-3.05,4.31,15.28) (0.00,9.38,18.46) (1.34,12.34,19.38)

99 Mohammad Ali Abdolv and Mohammad Javad Taghipouryan
Table 7: Service quality measures of customs by Fuzzy Servqual - continued
Responsiveness (-26.29,-14.56,-4.09) (-29.03,-18.33,-7.93) (-26.85,-12.83,-8.18)
A 14 (-15.28,-4.31,6.38) (-9.34,0.00,8.34) (-12.34,-3.45,6.28)
A 15 (0.00,0.56,16.28) (-4.31,1.25,10.25) (-1.35,7.36,19.28)
A 16 (14.35,25.38,34.29) (10.28,23.15,34.76) (6.38,15.28,29.28)
A 17 (-58.38,-43.29,-36.19) (-67.34,-49.31,34.19) (-64.31,-49.13,-36.19)
Assurance (-14.82,-5.41,5.19) (-17.67,-6.22,4.79) (-17.90,2.97,4.66)
E 18 (-12.34,-3.64,5.19) (-9.37,-1.06,7.36) (-10.54,-3.24,5.19)
E 19 (0.00,9.24,19.34) (3.64,18.61,28.16) (2.58,10.36,21.37)
E 20 (-8.24,0.00,6.34) (-4.25,3.45,12.58) (-3.16, 5.46,14.39)
E21 (-2.05,9.87,18.34) (0.00,8.42,17.34) (-2.64,7.38,19.34)
E 22 (-5.28,2.38,12.64) (-1.34,7.34,18.34) (2.64,11.34,21.54)
Empathy (-5.58,3.57,12.37) (-2.26,7.35,16.75) (-2.22,6.26,16.36)
Table 8: Overall service quality measures of customs
Twenty two criteria
T1 -48.41 -37.67 -29.33
T2 -1.20 -6.70 -12.31
T3 1.8 1.45 3.65
T4 -10.31 -15.79 -24.78
Tangibility -14.52 -14.67 -15.68
Rel 5 -47.27 -43.13 -47.70
Rel 6 27.35 10.79 4.11
Rel 7 -39.40 -39.41 -46.42
Rel 8 -50.75 -39.19 -53.28
Rel 9 -13.41 -2.86 -9.15
Reliability -27.81 -21.77 -30.56
Res 10 -49.65 -57.92 -57.32
Res 11 -14.26 -23.42 -24.30
Res 12 -5.70 -1.67 8.74
Res 13 2.64 9.28 11.02
Responsiveness -14.98 -18.43 -15.95
A 14 -4.40 -0.33 -0.87
A 15 5.61 2.40 8.43
A 16 24.67 2.17 4.22
A 17 -45.95 -27.49 -49.88
Assurance -5.01 -6.36 -3.42
E 18 -3.6 -1.02 -2.86
E 19 9.53 9.52 14.89
E 20 0.63 3.93 5.56
E21 8.75 8.29 8.02
E 22 3.25 8.11 6.56
Empathy 3.45 7.28 6.8
4.3. Final Ranking of Customs
This section applies TOPSIS to rank customs. For this purpose, we were obtaining the matrix of
performance to evaluate the three customs' performance by Servqual (table 8) and criteria weight by
Entropy (table.6). From table (9) to table (13) shows six steps of Topsis. As shown, final ranking of
customs are C, B. A.
 step 1 :
Table 9: Performance matrix
T Rel Res A E
A -14.52 -27.81 -14.98 -5.01 3.45
B -14.67 -21.77 -18.43 -6.36 7.28
C -15.68 -30.56 -15.95 -3.42 6.8

Evaluation of Customs Service Quality by Using Fuzzy SERVQUAL and Fuzzy MCDM 100
Table 10: Normalized performance matrix
T Rel Res A E
A -0.560 -0.595 -0.523 -0.570 0.327
B -0.565 -0.466 -0.644 -0.724 0.690
C -0.604 -0.654 -0.557 -0.389 -0.645
w 0.263 0.205 0.281 0.258 0.211
 Step 2:
Table 11: Weighted normalized performance matrix
T Rel Res A E
A -0.147 -0.121 -0.146 -0.147 0.068
B -0.148 -0.095 -0.180 -0.186 0.145
C -0.158 -0.134 -0.156 -0.100 0.136
 Step3: Determine the ideal solution and negative ideal solution
A+i = {-0.147, -0.095,-0.146,-0.100, 0.136}
A-i = {-0.158,-0.134,-0.180,-0.186,-0.068}
 Step 4:
Table 12: Distance between idea solution and negative ideal solution
A B C
S+ 0.116 0.132 0.041
S- 0.053 0.086 0.133
 Step5-6:
Table 13: Final ranking of instructors
Branch Similarity to ideal solution(C+) Rank
A 0.313 3
B 0.394 2
C 0.764 1
5. Conclusions and Implications
Service quality has been revealed as a key factor in search for sustainable competitive advantage,
differentiation and excellence in the service sector. Customers’ evaluations of the service quality are
critical to service firms that aim to improve their marketing strategies so Accurate measurement of
service quality is a major concern to management. But, whereas measuring criteria of services quality
and satisfaction are fuzzy and ambiguous but available methods measuring them generally is classic
kind. So Appling fuzzy methods according to other methods are closet to human thinking.
Therefore ¸ purpose of this paper is service quality evaluation of service organizations by using
Fuzzy MCDM approach.
In this research, Questionnaire that applied is extracted from criteria of service quality
(SERVQUAL). The study is presentation of a way for evaluation of Service Quality by using fuzzy
numbers on the basis of five dimensions of service quality in SERVQUAL model which we call Fuzzy
SERVQUAL.

101 Mohammad Ali Abdolv and Mohammad Javad Taghipouryan
For determining reliability of this questionnaire from in this research Cronbach alpha has been
used that alpha coefficients for the five dimensions were ideal Cronbach alpha (higher than 0.7).
We have chosen customs of Mazandaran province (Nowshar, Sari, and Amir Abad customs)
and 210 questionnaires distributed between clients (exporters, importers, commission agents and…)
that 194 of them were completed.
By applying Entropy, we obtained the weights of the five dimensions and twenty two of service
quality. The weights priorities for each of the dimensions are: Responsiveness (w = 0.281); Tangibility
(w = 0.263); Assurance (w = 0.258); Empathy (w = 0.211) and Reliability (w =0.205) (table.1).
After obtaining the criteria weights from Entropy, by using Fuzzy SERVQUAL evaluated
Service Quality of customs (Table 7). Then, we used Center-of- Area method to defuzzify the Fuzzy
numbers so as to conduct TOPSIS ranking procedure (Table 8). Finally, for ranking branches, Topsis
applied that customs final ranking from service quality are equal so as "custom C, custom B, custom
A" The concept of linguistic values and fuzzy numbers are used in this study since they could
easily be used to describe the subjective measurement of the appropriateness of alternatives and the
importance weightings of criteria.Fuzzy MCDM approach that used in this research, could use and help
for all of service organization that want to rank service organizations, for example : ranking banks
,hotels and so on.
References
[1] Agus, A., Barker, S. and Kandampully, J. (2007), “An exploratory study of service quality in
the Malaysian public service sector”, International Journal of Quality & Reliability
Management, Vol. 24 No. 2, pp. 177-90.
[2] Beheshti Hooshang M. & Lollar James G. (2008)" Fuzzy logic and performance evaluation:
discussion and application" International Journal of Productivity and Performance
Management, Vol. 57 No. 3, pp. 237-246.
[3] Brown, S.W. and Swartz, T.A. (1989), “A gap analysis of professional service quality”, Journal
of Marketing, Vol. 53, April, pp. 92-100.
[4] Brysland, A. and Curry, A. (2001), “Service improvements in public services using
SERVQUAL”, Managing Service Quality, Vol. 11 No. 6, pp. 389-401.
[5] Chamodrakas I. , Alexopoulou N., Martakos D. (2009)"Customer evaluation for order
acceptance using a novel class of fuzzy methods based on TOPSIS" Expert Systems with
Applications ,36, pp7409–7415.
[6] Carman, J.M. (1990), “Customer perceptions of service quality: an assessment of the
SERVQUAL dimensions”, Journal of Retailing, Vol. 66 No. 1, pp. 33-56.
[7] Chau V. S. & Kao Y.Y. (2009)"Bridge over troubled water or long and winding road? Gap-5 in
airline service quality performance measures" Managing Service Quality, Vol. 19 No. 1, pp.
106-134.
[8] Dubis, D. and Prade, H. (1978) Operations on fuzzy numbers. International Journal of Systems
Science, 9(7), 613-626.
[9] E. Abbott, James (1996) "Measuring quality with fuzzy logic" The tourist magazine, vol., no4
pp36-39.
[10] Fick, G.R. and Ritchie, J.R.B. (1991), “Measuring service quality in the travel and tourism
industry”, Journal of Travel Research, Vol. 30 No. 2, pp. 2-9.
[11] Ford, J.W., Joseph, M. and Joseph, B. (1993), “Service quality in higher education: a
comparison of universities in the United States and New Zealand using SERVQUAL”,
unpublished manuscript, Old Dominion University, Norfolk, VA.
[12] Freeman, K.D. and Dart, J. (1993), “Measuring the perceived quality of professional business
services”, Journal of Professional Services Marketing, Vol. 9 No. 1, pp. 27-47.

Evaluation of Customs Service Quality by Using Fuzzy SERVQUAL and Fuzzy MCDM 102
[13] Ghobadian, A., Speller, S. and Jones, M. (1994), “Service quality: concepts and models”,
International Journal of Quality & Reliability Management, Vol. 11 No. 9, pp. 43-66.
[14] Hoxley, M. (2000), “Measuring UK construction professional service quality: the what, how,
when and who”, International Journal of Quality & Reliability Management, Vol. 17 No. 4/5,
pp. 511-26.
[15] Jamali R. and Sayyadi H. (2009)"Prioritizing academic library service quality indicators using
fuzzy approach Case study: libraries of Ferdowsi University" Library Management, Vol. 30 No.
4/5,pp. 319-333.
[16] Johns, N. (1993), “Quality management in the hospitality industry. Part 3: Recent
developments”, International Journal of Contemporary Hospitality Management, Vol. 5 No. 1,
pp. 10-15.
[17] Kou Y. & Xiong J. (2006) "Entropy-based Improved TOPSIS as a Measure of Selecting Partner
in Strategic Alliance" International Journal of Business and Management, August, pp109-113.
[18] Kumar M.& Kee F.T. & Charles V. (2010)" Comparative evaluation of critical factors in
delivering service quality of banks: An application of dominance analysis in modified
SERVQUAL model" International Journal of Quality & Reliability Management, Vol. 27 No.
3,pp. 351-377.
[19] Kung, C.Y., Yan, T.M. and Lai, C.S. (2009), “Analyzing service quality in the mobile
communications industry: a comparison between GRA and LISREL methods”, Journal of Grey
System, Vol. 12 No. 1, pp. 49-58.
[20] Kasper, H., Helsdingen, P. and Gabbott, M. (2006), Services Marketing Management: A
Strategic Perspective, 2nd ed., John Wiley & Sons, Chichester.
[21] Liou, Tian-Shy & Chen,Ching-Wen (2006)" Subjective appraisal of service quality using fuzzy
linguistic assessment" International Journal of Quality & Reliability Management Vol. 23 No.
8.
[22] Lai W.-H.and Chang P.-L. And Chou Y.-C. (2010)" Fuzzy MCDM approach to R&D project
evaluation in Taiwan’s public sectors" Journal of Technology Management in China, Vol. 5
No. 1,pp. 84-101
[23] Langkumaran M.I and Kumanan S. (2009)” Selection of maintenance policy for textile industry
using hybrid multi-criteria decision making approach” Journal of Manufacturing Technology
Management, Vol. 20 No. 7, pp. 1009-1022
[24] Parasuraman, A., Berry, L.L. and Zeithaml, V.A. (1988), “SERVQUAL: a multiple- item scale
for measuring consumer perceptions of service quality”, Journal of Retailing, Vol. 64 No. 1, pp.
12-40.
[25] Rosenbaum M. S. & Wong I. A. (2009)"Modeling customer equity, SERVQUAL, and
ethnocentrism: a Vietnamese case study" Journal of Service Management, Vol. 20 No. 5, pp.
544-560.
[26] Sanayei A., Mousavi S. F., and Shahmirzadi C. A. (2009) "A Group Based Fuzzy MCDM for
Selecting Knowledge Portal System" World Academy of Science, Engineering and Technology
52 2009.
[27] Saleh, F. and Ryan, C. (1992), “Analysing service quality in the hospitality industry using the
SERVQUAL model”, Service Industries Journal, Vol. 11 No. 3, pp. 324-43.
[28] Selc¸uk Perc¸in(2010)" Use of analytic network process in selecting knowledge management
strategies" Management Research Review,Vol. 33 No. 5,pp. 452- 471.
[29] Shanian, A. & Savadogo, O. (2006) "TOPSIS multiple-criteria decision support analysis for
material selection of metallic bipolar plates for polymer electrolyte fuel cell" Journal of Power
Sources, 159, pp 1095–1104.
[30] Soo H. Y. (2004)" Towards the Development of a Decision Support System for Emergency
Vehicle Preemption and Transit Signal Priority Investment Planning" Dissertation submitted to

103 Mohammad Ali Abdolv and Mohammad Javad Taghipouryan
the Faculty of the Virginia Polytechnic Institute and State University, Doctor of Philosophy in
Civil and Environmental Engineering, April.
[31] Taghipouryan Y. & Taghipouryan M.J. and Nassimi M.A. (2010) "Gender Differences in
Service Quality Expectations in Iran Services Industries International Hotels-
Banks)"international conference on arts, sciences &technology, Malaysia,pp1-12.
[32] Zeithaml, V.A. and Parasuraman, A. (2004), Service Quality, Marketing Science Institute,
Cambridge, MA.
[33] Zeithaml, V.A., Parasuraman, A. and Berry, L.L. (1990), Delivering Quality Service: Balancing
Customer Perceptions and Expectations, The Free Press, New York, NY.