Download as pdf or txt
Download as pdf or txt
You are on page 1of 19

Int. J. Shipping and Transport Logistics, Vol. 6, No.

1, 2014 69

A comprehensive evaluation of the logistics service


quality based on vague sets theory

Ai-Hua Wu*
School of Business,
Ludong University,
186 Hong Qi Road,
Yantai 264025, China
E-mail: hiwuaihua@163.com
*Corresponding author

Jing-Qin Su
Faculty of Management and Economics,
Dalian University of Technology,
2 Ling Gong Road, Dalian 116024, China
E-mail: jingqin@dlut.edu.cn

Fei Wang
Glorious Sun School of Business and Management,
Donghua University,
1882 Yan An Xi Road, Shanghai 200051, China
E-mail: haomiaoshuitian@163.com

Abstract: The main purpose of the study is to establish a method for evaluating
the logistics service quality. Some studies have applied the SERVQUAL to
evaluate service quality, however most of these studies focus on the choice of
index, and the evaluation method of logistics service quality receives little
attention, especially fuzzy evaluation method. As a fact human judgments are
often vague and it is not easy to express the weights of evaluation criteria and
the satisfaction of logistics service quality with an exact numerical value. Due
to the existing fuzziness in the logistics service quality evaluation and a vague
set can provide more information than a fuzzy set, vague set theory is an
appropriate method for dealing with uncertainty. This article attempts to fill this
gap in the current literature by establishing an improved fuzzy comprehensive
evaluation method based on vague sets for evaluating the logistics service
quality. The method may reflect both for/true and against/false judgment
information in practice and it will efficiently manage the essential fuzziness in
human judgment and preference. Moreover, an example of the evaluation
process of logistics service quality is shown to illustrate the proposed theory
and method.

Keywords: logistics service; quality evaluation; vague sets; SERVQUAL


method; probability matrix.

Copyright © 2014 Inderscience Enterprises Ltd.


70 A-H. Wu et al.

Reference to this paper should be made as follows: Wu, A-H., Su, J-Q. and
Wang, F. (2014) ‘A comprehensive evaluation of the logistics service quality
based on vague sets theory’, Int. J. Shipping and Transport Logistics, Vol. 6,
No. 1, pp.69–87.

Biographical notes: Ai-Hua Wu is an Assistant Professor in the School


of Business at the Ludong University. She earned her PhD in Business
Administration from the Glorious Sun School of Business and Management at
the Donghua University. Her research interests include transports, logistics
management, technological innovation and human capital. She has authored or
co-authored three books, some number of journal articles and international
conference papers.

Jing-Qin Su is a Professor of the Faculty of Management and Economics at the


Dalian University of Technology. He earned his doctoral degree from Dalian
University of Technology. His recent research interests include supply chain
management, corporate strategies and technological innovation. His studies
have been published in several top refereed journals.

Fei Wang is a doctoral candidate of the Glorious Sun School of Business and
Management at the Donghua University. Her current research interests include
supply chain management, fuzzy and vague sets. She has published some
number of journal articles and international conference papers.

1 Introduction

With the e-business booming across the world in the last decade, logistics industry has
developed rapidly. Logistics and supply chain management (SCM) literatures indicate
that customer service management has become one of the most important strategic issues
for companies in recent years. By improving logistics performances, companies can
increase customer satisfaction and gain market shares. The role of SCM in creating and
maintaining a strategic competitive advantage has been recognised through increased
customer value and satisfaction and the attendant business profitability (Mentzer et al.,
2001). An important aspect of SCM is consistent, high-quality logistics services. With the
pressure of competition, many service industries have decided to improve service quality
and service items in order to differentiate their services from those of their competitors
(Kuo et al., 2007). So understanding, building and maintaining as well as evaluating
service quality are the main concerns of business today.
It is generally believed that logistics service quality (LSQ) should include technical
quality and functional quality, and there are a great number of literatures about the choice
of evaluation index, such as the famous SERVQUAL instrument, which consists of five
dimensions of service quality: tangibility, reliability, responsiveness, assurance and
empathy proposed by Parasuraman and Zeithaml as well as Berry (Parasuraman et al.,
1988). A number of studies have applied the SERVQUAL to evaluate the LSQ in the
past, however most of these studies focus on the choice of index for LSQ evaluation to
specific logistics industry, and the evaluation method receives little attention. As a fact,
logistics service industry contains intangibility, inseparability and heterogeneity, and the
evaluation criteria often simultaneously have independent, dependent and interdependent
characteristics (Kuo et al., 2007). So the evaluation of LSQ has become very complex
A comprehensive evaluation of the logistics service quality 71

and difficult in a real-life environment, because the result defends on human judgments
and preference, which are often vague and cannot estimate his/her preference with an
exact numerical value.
Bellman and Zadeh (1970) and Hsu and Chen (1997) consider that it is more realistic
to use linguistic terms to describe the desired value and important weight of criteria, e.g.,
‘very low’, ‘low’, ‘fair’, ‘high’, ‘very high’, etc. Due to this type of existing fuzziness in
the process, fuzzy set (FS) theory is an appropriate method for dealing with uncertainty,
and it was introduced by Zadeh (1965) in 1965 to solve problems involving the absence
of sharply defined criteria. The subjective evaluation data can be more adequately
expressed in linguistic variables (Chen, 2000; Liang and Wang, 1991). Because fuzziness
and vagueness are common characteristics in many decision-making problems, especially
in service quality evaluation, good evaluation models should be able to tolerate
vagueness or ambiguity (Yang et al., 2012), and a major contribution of FS theory is its
capability of representing vague data. Thus the method of fuzzy comprehensive
evaluation based on FS theory is a common method in evaluation of industry and
management area.
However in FS theory, each object u ∈ U (Here U is a classical set of objects, called
the universe of discourse.) is assigned a single real value, called the grade of
membership, between zero and one. Gau and Buehrer (1993) point out that the drawback
of using the single membership value in FS theory is that the evidence for u ∈ U and the
evidence against u ∈ U are in fact mixed together. In order to overcome this drawback,
they propose vague set (VS) theory and define two values: the true and false membership
functions, so it allows using interval-based membership instead of using point-based
membership as in a FS. The interval-based membership generalisation in VSs is more
expressive in capturing vagueness of data, which expands the FS theory and can express
more information including fuzzy and uncertainty information. Since VSs can provide
more information than FSs, they are considered superior in mathematical analysis of
uncertain information.
In this paper, an improved fuzzy comprehensive evaluation method based on VSs is
proposed and applied to evaluate LSQ from the perspective of customer, which may
reflect both for/true and against/false judgment information in human judgment and
preference. And an example of the evaluation process of LSQ is shown to illustrate the
proposed theory and method and highlight the procedure of the proposed approach,
which will be useful for users choosing logistics service supplier.
The remainder of this paper is organised as follows. Section 2 reviews LSQ and
methods for evaluating service quality. Section 3 describes the concepts of FSs and VSs
and the analysis process. In Sections 4 and 5, the proposed method is illustrated with an
example of the evaluation process of LSQ. Finally, conclusions are presented in
Section 6.

2 Literature review

2.1 Logistics service quality


Logistics, that is, the management of business operations related to the acquisition,
storage, transportation and delivery of goods along the supply chain, is one of the basic
business functions (Irene and Maria, 2011). In the service marketing literature,
72 A-H. Wu et al.

services are frequently described by characteristics such as intangibility, heterogeneity,


inseparability of consumption from production and the impossibility to keep services in
stock. Service quality can be regarded as a composite of various attributes. It not only
consists of tangible attributes, but also intangible and subjective attributes such as safety,
comfort and satisfaction, which are difficult to measure accurately (Chou et al., 2011).
As for LSQ, Perreault and Russ (1974, 1976) consider that logistics activities create
time, place, and form utility, thereby enhancing product value. This assertion was
extended by some scholars (e.g., Coyle et al., 1992; Shapiro and Heskett, 1985). Stock
and Lambert (1987) describe the seven Rs of utility creation by logistics services:
delivering the right amount, the right product, at the right place, in the right condition, at
the right time, with the right information, and at the right price. Continuing the
development of LSQ research, Mentzer et al. (1989) assert that delivery service quality
consists of ‘customer service quality’ and ‘physical distribution service quality’. The
marketing customer service component proposed by Mentzer et al. (1989) echoes a
contention by Perreault and Russ (1974), who maintain that understanding perceptions of
physical distribution service from a customer’s point of view is an essential input in
marketing management decisions.
So the research about LSQ can be found in the marketing service quality literature.
Grönroos (1983, 1984) develops the concept of perceived service quality, two quality
dimensions are called functional quality (how the service process functions) and technical
quality (what the service process leads to for the customer in a ‘technical’ sense) in the
perceived service quality model. Zeithaml (1988) discusses the concept of perceived
versus objective quality and considers that perceived quality is the consumer’s judgment
about superiority or excellence. As an integration of the logistics and marketing
service quality research streams, Parasuraman et al. (1988) develop a multiple-item scale
(SERVQUAL) for measuring service quality and simplify the ten dimensions of service
quality to five dimensions. The ten dimensions are reliability, responsiveness,
competence, access, courtesy, communication, credibility, security, understanding and
knowing the customer and tangibles; and the five dimensions are tangibles, reliability,
responsiveness, assurance and empathy. And they make further evaluations and
improvements to SERVQUAL to apply it to different types of service industries. The
SERVQUAL method is one of the best evaluation methods for assessing the expectations
and perceptions, although there are some difficulties encountered in applying it to an
industrial service context (Bienstock et al., 1997).

2.2 Methods for evaluating the service quality


Several fuzzy methods for evaluating the service quality have been proposed and applied
to solve practical problems. Fuzzy analytic hierarchy process (FAHP) approach is a
widely used method. For example, Mikhailov and Tsvetinov (2004) use it for tackling the
uncertainty and imprecision of the service evaluation process; Ozcan and Suzan (2011)
use it to select the best supplier firm providing the most customer satisfaction for the
criteria determined. A novel FAHP approach for addressing uncertainty and imprecision
in service evaluation during pre-negotiation stages is proposed by Chang et al. (2008),
and comparative judgments of decision makers (DMs) are represented as triangular fuzzy
numbers by them. Conventionally designed questionnaire frequently use the Likert scale
to gauge the feeling of respondents. Owing to the fuzziness of human thinking, Chou
(2009) considers that this approach is inadequate and too simple to rule subject’s way and
A comprehensive evaluation of the logistics service quality 73

measure complex human thinking and cognition. A fuzzy SERVQUAL approach to


clarify the positioning of service quality in the department store market and
implementation priorities for different service strategies is proposed by Tsai et al. (2008).
Chou (2009) uses the fuzzy weighted SERVQUAL method in the evaluation of
airline service quality, and a case study of Taiwanese airline is conduced to demonstrate
the effectiveness of the fuzzy weighted SERVQUAL model. Fuzzy multiple criteria
decision-making (FMCDM) and fuzzy multiple attribute decision-making (FMADM)
approaches have been proposed and applied to the evaluation of service quality (Chou,
2009; Hu, 2009; Shih and Hu, 2008). Fuzzy quality function deployment (FQFD) models
have been also introduced and used to solve service quality evaluation problems (Liang
et al., 2006; Sultan and Simpson, 2000). Xu (2008) uses gray correlation method to
analyse LSQ evaluation outcome. In addition, some integrated algorithms also are
applied to the evaluation of service quality (Montana and Hussain, 2004; Chandramathi
et al., 2008).

3 Fuzzy, VSs and intuitionistic fuzzy sets

3.1 Fuzzy sets


The FS theory was introduced by Zadeh (1965). Fuzzy numbers are a fuzzy subset of real
numbers, and they present the expansion of the idea of confidence interval. FS theory has
been developed exactly based on the premise that the key elements in human thinking are
not numbers, but linguistic terms or labels of FSs (Zimmermann, 1991). FS theory has
been applied to solve many decision-making problems. A major contribution of FS
theory is its capability of representing vague data. Because fuzziness and vagueness are
common characteristics in many decision-making problems, if the uncertainty (fuzziness)
of human decision making is not taken into account, the results from the models can be
misleading. A FS is a class of objects with a continuum of grades of membership. Such a
set is characterised by a membership function, which assigns to each object a grade of
membership ranging between zero and one.
Let U be the universe of discourse which is a classical set of objects, in FS theory,
each object u ∈ U is assigned a single real value, called the grade of membership,
between zero and one. Gau and Buehrer (1993) point out that the drawback of using the
single membership value in FS theory is that the evidence for u ∈ U and the evidence
against it is in fact mixed together. In order to tackle this problem, Gau and Buehrer
(1993) propose the notion of VSs, which allow using interval-based membership
instead of using point-based membership as in FSs. The interval-based membership
generalisation in VSs is more expressive in capturing vagueness of data.

3.2 Vague sets


A VS is a set of objects, each of which has a grade of membership whose value is a
continuous subinterval of [0, 1]. Such a set is characterised by a truth-membership
function and a false-membership function, and the true and false membership
function form the bound of the membership functions μA(u), which means that
tA(u) ≤ μA(u) ≤ 1 – fA(u). The main characteristic of VS is that it can offer for/true and
against/false evidence at the same time. For example, [tA(u), 1 – fA(u)] = [0.5, 0.8], then
74 A-H. Wu et al.

tA = 0.5, 1 – fA = 0.8, and fA = 0.2, i.e., the degree of acceptance is 0.5 and degree of
rejection is 0.2.
Let U be the universe of discourse, U = {U1 , U 2 , U n }, with a generic element of U
denoted by ui. The VS A from U can be expressed by a true membership function tA and a
false membership function fA, and tA(u) is the lower bound derived from evidence
supporting U, fA(u) is the lower bound derived from evidence against U, then tA(u)
and fA(u) make links between one of the interval [0, 1] and each point in U, namely:
U → [0, 1] and fA: U → [0, 1]. It is easy to know that tA(u) + fA(u) ≤ 1. The membership
of element u in VS A can be defined by subinterval [tA(u), fA(u)] of interval [0, 1], the
subinterval is vague value of element u in A, which is noted by vA(u).
When the universe of discourse U is continuous, a VS A can be written as

A= ∫ [t
U
A (u ), 1 − f A (u ) ] u = ∫
U
v A (u ) u (u ∈ U )

When the universe of discourse U is discrete, a VS A can be written as


n n
A= ∑i =1
⎡⎣t A ( ui ) , 1 − f A ( ui )⎤⎦ ui = ∑v
i =1
A ( ui ) ui (u ∈ U )

The uncertain information about u can be expressed by 1 – fA(u) – tA(u), and if it is


smaller, we know more accurately about u, on the contrary, if it is bigger, we know less
information about u. If 1 – fA(u) = tA(u), we know u accurately, and the VS is a FS. Let A
be a VS of the universe of discourse U. If ∀i, 1 ≤ i ≤ n, tA(ui) = 1 and fA(ui) = 0, then A is
called a unit VS; If ∀i, 1 ≤ i ≤ n, tA(ui) = 0 and fA(ui) = 1, then A is called a zero VS.

3.2.1 Operation between VSs


Let A and B be two VSs of the universe of discourse U. If ∀i, 1 ≤ i ≤ n, tA(ui) = tB(ui);
fA(ui) = fB(ui), then the VSs A and B are called equal; If ∀i, 1 ≤ i ≤ n, tA(ui) ≤ tB(ui) and
1 – fA(ui) ≤ 1 – fB(ui), then the VS A is included by B, denoted by A ⊆ B.
The intersection of the VSs A and B is a VS C, written as C = A ∧ B, whose truth
membership function and false-membership function are tC and fC respectively, where
∀ui ∈ U,

tC ( ui ) = t A ( ui ) ∧ t B ( ui ) = Min ( t A ( ui ) , t B ( ui ) )

1 − fC ( ui ) = (1 − f A ( ui ) ) ∧ (1 − f B ( ui ) ) = Min (1 − f A ( ui ) , 1 − f B ( ui ) )

The union of the VSs A and B is a VS D, written as D = A ∨ B, whose truth membership


function and false-membership function are tD and fD respectively, where ∀ui ∈ U,

t D ( ui ) = t A ( ui ) ∨ t B ( ui ) = Max ( t A ( ui ) , t B ( ui ) )

1 − f D ( ui ) = (1 − f A ( ui ) ) ∨ (1 − f B ( ui ) ) = Max (1 − f A ( ui ) , 1 − f B ( ui ) )

The complementary of VS A is a VS Ac, whose truth-membership function and false


membership function are t AC and f AC respectively, where ∀ui ∈ U,

t AC ( ui ) = f A ( ui )
A comprehensive evaluation of the logistics service quality 75

1 − f AC ( ui ) = 1 − t A ( ui )

3.2.2 Comparison between VSs


Let vague values x = [tx, 1 – fx], y = [ty, 1 – fy], where tx, fx, ty, fy ∈ [0, 1], the possibility of
x ≥ y is defined by
Max ( 0, Lx + Ly − Max ( 0, 1 − f y − t x ) )
P( x ≥ y) =
Lx + Ly

where Lx = 1 – fx – tx and Ly = 1 – fy – ty are the length of vague values x and y,


respectively. So
1 0 ≤ P(x ≥ y) ≤ 1
2 if P(x ≥ y) = P(y ≥ x), then P(x ≥ y) = P(y ≥ x) = 0.5
3 P(x ≥ y) + P(y ≥ x) = 1
4 for any three vague values x, y and z, if P(x ≥ y) ≥ 0.5, P(y ≥ z) ≥ 0.5, then
P(x ≥ z) ≥ 0.5.
In addition, Let A, B be two VSs in the universe of discourse U = {u1 , u2 , un },

∑ ∑
n n
A= [t A (ui ), 1 − f A (ui )] / ui , B = [t B (ui ), 1 − f B (ui )] / ui ; the probability of A ≥ B
i =1 i =1
is defined by
n
1
P ( A, B ) =
n ∑P(μ
i =1
A ( ui ) ≥ μB ( ui ) ) .

3.2.3 Similarity measure between VSs


Zhang et al. (2004) propose a method to calculate the similarity measure between vague
values x = [tx, 1 – fx] and y = [ty, 1 – fy]:
d ( x, y )
S ( x, y ) = 1 −
2

where d ( x, y ) = (t x − t y ) 2 + (1 − f x − (1 − f y )) 2 = (t x − t y ) 2 + ( f x − f y ) 2 is the distance


between x and y. Also, the similarity measure between VSs can be defined as follows:
Let A, B be two VSs in the universe of discourse U = {u1 , u2 , un },

∑ ∑
n n
A= [t A (ui ), 1 − f A (ui )] / ui , B = [t B (ui ), 1 − f B (ui )] / ui ; the similarity measure
i =1 i =1
between A and B is
n
1
S ( A, B ) =
n ∑S (μ
i =1
A ( ui ) , μB ( ui ) ) .
76 A-H. Wu et al.

3.2.4 Defuzzification of vague values


For x = [tx, 1 – fx], define the defuzzification function to get the precise value as follows:
Dfzz ( x) = t x ( t x + f x ) .

3.2.5 Weighted sum of vague values


For xi = [t xi , 1 − f xi ], whose weights vector w = {w1 , w2 , wn } are n precise values, the
weighted sum of xi (i = 1,…,n) is defined as follows:
n
⎡ n n

x= ∑
i =1

wi × xi = ⎢ wi × ti , 1 −
⎣⎢ i =1
∑ w × f ⎥⎦⎥
i =1
i i

n
where ∑ w = 1.
i =1
i

3.3 VSs and IFSs


Besides VS, intuitionistic fuzzy set (IFS) is also a further generalisation of FS. IFSs
theory was introduced by Atanassov (1986) in 1986, which can process the information
of membership and non-membership and provide more choices for the attribute
description of an object. So IFS has stronger ability to express uncertainty than Zadeh’s
FS.
An IFS { < u, μA(u), vA(u) > | u ∈ U} in a universe of discourse U is characterised by
a membership function μA, and a non-membership function vA, as follows: μA: U → [0, 1],
vA: U → [0, 1], and (μA + vA) ∈ [0, 1].

Figure 1 Membership functions of a VS


Membership
Functions Against Region
1

Hesitation Region
1 − f A (u )
1 − f A (u 0 )

t A (u0 )

Support Region t A (u )
O u0
U
A comprehensive evaluation of the logistics service quality 77

Figure 2 Membership functions of an IFS


Membership
Functions
1

μA (u )

vA ( u0 )

μA ( u0 ) vA (u)

O u0
U

It is easy to know that the difference between VSs and IFSs is due to the definition of
membership intervals (An and Wilfred, 2005). The subtle difference gives rise to a
simpler but meaningful graphical view of datasets, depicting a VS in Figure 1 and an IFS
in Figure 2 respectively. It can be seen that, the shaded part formed by the boundary in a
given VS in Figure 1 naturally represents the possible existence of data. Thus, this
‘hesitation region’ corresponds to the intuition of representing vague data. There are
more benefits of using vague membership intervals in capturing data semantics (An and
Wilfred, 2005).

4 Formalising the methodology

The amount of candidate schemes is K, and A = { A1 , A2 , Ak } is the set of K,


C = {C1 , C2 , C N } is the evaluation index. W = {W1 , W2 , WN } is the index weights of
C = {C1 , C2 , C N }. Assume that a decision group has K persons, weights vector
D = {d1 , d 2 , d n } can be obtained by professional knowledge and experience of experts,
which is the subjective weights vector of DMs. Dk (k = 1,…,K) is the importance degree
of the kth DM, and Dk = [t Dk , 1 − f Dk ] is a linguistic variable in Table 1. In addition, the
evaluation weights and evaluation view of schemes are linguistic variables, and the
linguistic variables can be expressed in vague numbers according to Table 1 (Zhang
et al., 2009).
Table 1 The scale of linguistic variables

Attribute ratings Weights Vague value


Very poor, VP Very low, VL [0.0, 0.1]
Poor, P Low, L [0.1,0.3]
Medium poor, MP Medium low, ML [0.3, 0.4]
78 A-H. Wu et al.

Table 1 The scale of linguistic variables (continued)

Attribute ratings Weights Vague value


Fair, F Medium, M [0.4, 0.5]
Medium good, MG Medium high, MH [0.5, 0.6]
Good, G High, H [0.6, 0.9]
Very good, VG Very high, VH [0.9, 1.0]

The selection processes are summarised as follows:


1 Use linguistic variables to evaluate the weights of each index, W k = {W1k ,W2k , Wnk }
(k = 1,…,K) is the vector of weights of index C = {C1 , C2 , Cn } given by kth DM,
where W jk = [twkj , 1 − f wkj ]( j = 1,… , n) is a linguistic variable in Table 1.

2 Use linguistic variables to judge for the evaluation views of schemes, and the
evaluation matrix given by kth DM is

⎡ R11k k
R12 R1kn ⎤
⎢ k k ⎥
R R22 R2kn ⎥
R k = ⎢ 21
⎢ ⎥
⎢ k ⎥
⎣⎢ Rm1 Rmk 2 k
Rmn ⎦⎥

where Rijk (i = 1, , m; j = 1, n) is the evaluation view of alternative scheme Ai on


index Cj given by kth DM, and Rijk = [t Rijk , 1 − f Rijk ] is also a linguistic variable in
Table 1.
3 Calculate the weighted decision matrix of kth DM. Considering the different
importance of each index, the weighted decision matrix can be expressed as

⎡ M 11k M 12k M 1kn ⎤


⎢ k k ⎥
M M 22 M 2kn ⎥
M k = ⎢ 21 (1)
⎢ ⎥
⎢ k ⎥
⎢⎣ M m1 M mk 2 k
M mn ⎥⎦

where
M ijk = W jk ∧ Rijk (i = 1,… , m; i = 1,… , n).


n
Each line M ik = M ijk Ci represents the evaluation of kth DM vis-a-vis
j =1

alternative Ai on index set C = {C1 , C2 , Cn }. It is also a VS.

4 Adjust the importance degree of DMs. The final decision should be close to the
preference of most DMs, so increasing the weight of DMs whose preference is close
to the group preference is reasonable. Calculate the similarity between the pth DM
and qth DM as follows:
A comprehensive evaluation of the logistics service quality 79

∑ S (M )
1
S pq = S ( M p , M q ) = p
i , M iq
m i =1

Then, it is easy to get the preference accordance matrix of all DMs:


⎡ S11 S12 S1K ⎤
⎢S S22 S2 K ⎥⎥
S = [ S pq ] = ⎢
21
(2)
⎢ ⎥
⎢ ⎥
⎣ SK1 SK 2 S KK ⎦
m
1
where S pq = S ( M p , M q ) =
m ∑ S (M
i =1
p
i , M iq ).

Obviously, S is a symmetry matrix. Using the line sum of S, the similarity weights
vector e = {e1 , e2 , e K }(k = 1,… , K ) is obtained as follows:

∑ ∑ S −1
K K
Skq kq
q =1, q ≠ k q =1
ek = = (3)
∑ ∑ ∑ ∑ S −K
K K K K
S pq pq
p =1 q =1, q ≠ k p =1 q =1

Since e is derived from the preference matrix given by all DMs, it is called the
objective weights vector.
5 Adjust the weights vectors of all DMs by both subjective and objective weights
vectors. The precise value w = {w1 , w2 , wK } of the subjective weights vector
D = {D1 , D2 , DK } is obtained as follows:

wk = t Dk ( tD k + f Dk ) (4)

Normalise w to get the final subjective weights vector w′. In this paper, a new
parameter, α is considered, calculate the adjusted weights vector d = {d1 , d 2 , d K }
as follows:
d k = α × wk′ + (1 − α ) × ek , k = 1, 2,… , K , (5)

where α ∈ [0, 1] represents the preference to subjective weights against objective


weights. The larger α is, the more is the attention of DMs to subjective weights.
When α = 0, it shows that the DM is the most objective, only considering the
objective weights. On the contrary, when α = 1, it shows that the DM is the most
subjective, only considering subjective weights. And when α = 0.5, it shows
that the DM is fair and subjective weights equivalent to objective weights.
6 Integrate all DMs’ preference matrix to generate the whole decision matrix
⎡ H11 H12 H1n ⎤
⎢H H 22 H 2 n ⎥⎥
H = ⎢ 21 (6)
⎢ ⎥
⎢ ⎥
⎣ H K1 HK2 H Kn ⎦
80 A-H. Wu et al.

where
K
⎡ K K

H ij = ∑
k =1 ⎢⎣ k =1

d k × M ijk = ⎢ d k × tM ijk , 1 − ∑d
k =1
k × f M ijk ⎥
⎥⎦

Each line Hi in matrix H represents the evaluation of alternative Ai from the whole
decision group. Obviously, Hi is a VS.
7 According to the formula of possibility degree to vague value, comparing the
evaluation values Hi and Hl, get the possibility matrix corresponding of all
alternatives
⎡ P11 P12 … P1m ⎤
⎢P P22 … P2 m ⎥⎥
P=⎢
21
(7)
⎢ ⎥
⎢ ⎥
⎣ Pm1 Pm 2 … Pmn ⎦
n
1
where Pil = P ( H i ≥ H l ) =
n ∑ P( H
j =1
ij ≥ H lj ).

8 Calculate the order vector of all alternatives. The sortings vector of candidate
schemes set A = { A1 , A2 , Ak } can be expressed by
K

∑p
l =1
jl

Oj = K K
(8)
∑∑ p
j =1 l =1
jl

when Oj is bigger, the sorting order of Aj is better.

5 Experimental results

Company G is a hi-tech enterprise of China, by focusing on smart grid, internet of things


(IOT), environmental protection and energy saving, it has provided the intelligent
automation control products and solutions in terms of power generation, power
transmission, power transformation, power distribution and power utilisation, etc. It has
come out top in the market share of China’s power automation industry for more than ten
years and exported the technologies and services to South America, Southeast Asia,
Middle East, Europe, etc. In the process of operating, it depends on logistics service
suppliers to ship its products to customers. For example, there are three logistics services
suppliers A1, A2, A3, and five indicators C1, C2, C3, C4, C5 are applied to evaluate
service quality based on SERVQUAL, which indicate reliability, responsiveness,
assurance, empathy and tangibles respectively. A committee of four DMs including
general manager (DM1), sales manager (DM2), office director (DM3) and a product line
manager (DM4) has been formed to make the selection decision, whose importance
weights are shown in Table 2. The evaluation frame is shown in Figure 3.
A comprehensive evaluation of the logistics service quality 81

Table 2 Decision makers’ importance weights

DM DM1 DM2 DM3 DM4


Weights VH VH H MH

Figure 3 Evaluation frame

Identify the attribute weights and attribute ratings of alternatives. The index weights
given by four DMs are shown in Table 3.
Table 3 Index weights

Item DM1 DM2 DM3 DM4


C1 VH VH VH VH
C2 H VH H VH
C3 MH H H H
C4 VH H MH MH
C5 M MH MH M

The index ratings of alternatives given by four DMs are shown in Table 4.
Table 4 The index evaluation view for all alternatives

Item DM DM1 DM2 DM3 DM4


C1 A1 MG MG G G
A2 MG G F MG
A3 VG VG MG G
82 A-H. Wu et al.

Table 4 The index evaluation view for all alternatives (continued)

Item DM DM1 DM2 DM3 DM4


C2 A1 G G MG MG
A2 G MG MG G
A3 F F F F
C3 A1 G MG MG G
A2 MG G G MG
A3 G G G MG
C4 A1 F G F G
A2 F G F MG
A3 F F G G
C5 A1 G MG G G
A2 G VG VG G
A3 MG MG G MG

Calculate the weighted decision matrix for all suppliers. According to formula (1), we
can obtain the weighted decision matrix as shown in Table 5.
Table 5 The weighted decision matrix

Item DM DM1 DM2 DM3 DM4


C1 A1 [0.5 0.6] [0.5 0.6] [0.6 0.9] [0.6 0.9]
A2 [0.5 0.6] [0.6 0.9] [0.4 0.5] [0.5 0.6]
A3 [0.9 1.0] [0.9 1.0] [0.5 0.6] [0.6 0.9]
C2 A1 [0.6 0.9] [0.6 0.9] [0.5 0.6] [0.5 0.6]
A2 [0.6 0.9] [0.5 0.6] [0.5 0.6] [0.6 0.9]
A3 [0.4 0.5] [0.4 0.5] [0.4 0.5] [0.4 0.5]
C3 A1 [0.5 0.6] [0.5 0.6] [0.5 0.6] [0.6 0.9]
A2 [0.5 0.6] [0.6 0.9] [0.6 0.9] [0.5 0.6]
A3 [0.5 0.6] [0.6 0.9] [0.6 0.9] [0.5 0.6]
C4 A1 [0.4 0.5] [0.6 0.9] [0.4 0.5] [0.5 0.6]
A2 [0.4 0.5] [0.6 0.9] [0.4 0.5] [0.5 0.6]
A3 [0.4 0.5] [0.4 0.5] [0.5 0.6] [0.5 0.6]
C5 A1 [0.4 0.5] [0.5 0.6] [0.5 0.6] [0.4 0.5]
A2 [0.4 0.5] [0.5 0.6] [0.5 0.6] [0.4 0.5]
A3 [0.4 0.5] [0.5 0.6] [0.5 0.6] [0.4 0.5]

Adjust the importance degree of DMs. According to formula (2), we can get the
preference accordance matrix of four DMs as follows:
⎡1.0000 0.9677 0.9737 0.9820 ⎤
⎢0.9677 1.0000 0.9547 0.9498 ⎥⎥
S=⎢
⎢0.9737 0.9547 1.0000 0.9780 ⎥
⎢ ⎥
⎣ 0.9820 0.9498 0.9780 1.0000 ⎦
A comprehensive evaluation of the logistics service quality 83

According to formula (3), the similarity weights vector of four DMs can be obtained as
follows:
e = {0.2518, 0.2476, 0.2503, 0.2506}

Adjust the weights vector of four DMs. According to formula (4), the precise value of the
subjective weights vector D is obtained as follows:
w = {1.0000, 1.0000, 0.8571, 0.5556}

Normalise w into w′:


w′ = {0.2930, 0.2930, 0.2512, 0.1628}

Assume α = 1 (subjective, only considering subjective weights), α = 0 (objective, only


considering the objective weights) and α = 0.5 (fair, subjective weights were assumed to
have the same importance as objective weights) respectively, according to formula (5),
we can get the adjusted weights vector as shown in Table 6.
Table 6 Adjusted weights vector

Item d1 d2 d3 d4
α=1 0.2930 0.2930 0.2512 0.1628
α=0 0.2518 0.2474 0.2503 0.2506
α = 0.5 0.2724 0.2702 0.2506 0.2068

Generate the whole decision matrix. According to formula (6), the integrated decision
matrix of four DMs are obtained as follows:
⎡[0.5707 0.8121] [0.5659 0.7978] [0.5871 0.8613] [0.5808 0.8423]⎤
H (α =1) = ⎢⎢[0.5042 0.6628] [0.5007 0.7269] [0.4719 0.6071] [0.4968 0.6514]⎥⎥
⎢⎣[0.4577 0.5903] [0.4891 0.6444] [0.5169 0.6927] [0.5169 0.7113]⎥⎦

⎡[0.5753 0.8258] [0.5686 0.8059] [0.5859 0.8578] [0.5825 0.8475] ⎤


H (α = 0) = ⎢⎢[0.4997 0.6492] [0.5004 0.6631] [0.4748 0.6029] [0.4937 0.6287]⎥⎥
⎢⎣ [0.4751 0.6253] [0.4936 0.6562] [0.5169 0.6946] [0.5217 0.7196]⎥⎦

⎡[0.5730 0.8189] [0.5672 0.8015] [0.5864 0.8593] [0.5817 0.8450]⎤


H (α = 0.5) = ⎢⎢[0.5020 0.6560] [0.5006 0.6663] [0.4734 0.6005] [0.4952 0.6318]⎥⎥
⎢⎣[0.4664 0.6078] [0.4909 0.6495] [0.5167 0.6933] [0.5193 0.7155] ⎥⎦

According to formula (7), we can obtain the possibility degree matrixes as follows:
⎡ 0.5000 0.7999 0.8060 ⎤
P(α =1) = ⎢⎢ 0.2001 0.5000 0.5008 ⎥⎥
⎢⎣ 0.1940 0.4992 0.5000 ⎥⎦
84 A-H. Wu et al.

⎡ 0.5000 0.8553 0.7796 ⎤


P(α = 0) = ⎢⎢ 0.1447 0.5000 0.4262 ⎥⎥
⎣⎢ 0.2204 0.5738 0.5000 ⎥⎦

⎡0.5000 0.8461 0.7929 ⎤


P(α = 0.5) = ⎢⎢0.1539 0.5000 0.4492 ⎥⎥
⎢⎣ 0.2071 0.5508 0.5000 ⎥⎦

According to formula (8), we can obtain the sortings vector as follows:


O(α =1) = {0.4680, 0.2668, 0.2652}

O(α = 0) = {0.4744, 0.2380, 0.2876}

O(α = 0.5) = {0.4753, 0.2451, 0.2795}

Therefore, in cases of α = 0.5 and α = 0, the rankings order of three logistics suppliers
will be as follows: A1 > A3 > A2, however when α = 1, which means only considering
subjective weights, the order is: A1 > A2 > A3. So we can say that the service of A1 is the
best among three logistics suppliers. A1 should be an important alternative for the
company.

6 Conclusions

Logistics is service in the essence, and LSQ evaluation is a multi-objective evaluation


which contains uncertain information. This research is based on the five dimensions of
SERVQUAL with the theory of VS to build an evaluation framework of LSQ. In the past,
many researchers evaluated the service quality using the SERVQUAL method, few of
these researchers applied the fuzzy theory to evaluate the LSQ. Most of previous
SERVQUAL questionnaires are based on accurate measure (Chou et al., 2011). Since
human judgments and preference are often vague and cannot estimate his/her preference
with an exact numerical value. Thus, VSs theory is an appropriate method, and it is a
novel mathematical tool for dealing with incomplete and imprecise information.
The major contribution of this paper lies in the development of an improved fuzzy
comprehensive method based on VSs, which incorporates diversified issues, for the
evaluation of LSQ under uncertainty environments. FS theory has been proposed to
handle such vagueness by generalising the notion of membership in a set. Essentially, in a
FS each element is associated with a point-value selected from the unit interval [0, 1],
which is termed the grade of membership in the set. A VS is a further generalisation of a
FS. Instead of using point-based membership as in FSs, interval-based membership is
used in a VS, thus interval-based membership in VSs is more expressive in capturing
vagueness of data.
In this paper, a new comprehensive evaluation method based on VSs theory is
proposed, which can make full use of the advantage of VSs in coping with uncertain
information. The step is shown briefly as follows: Firstly, use linguistic values to assess
the ratings and weights for quantitative or qualitative factors. Secondly, define the
concept of similarity and possibility degree, by establishing the probability matrix get the
A comprehensive evaluation of the logistics service quality 85

ranking vector of the alternatives. Finally, an example of the evaluation process of LSQ is
shown to highlight the procedure of the proposed approach, and the results show that the
proposed method can obtain satisfactory results. At a time when outsourcing of logistics
activities has become a global trend, this paper provides an insight into the choice of
logistics supplier. Although the proposed method presented in this paper is illustrated by
a selection process for criteria of service quality, it can also be applied to other selection
processes.
The main aim of the paper is to propose a new VS-based approach to evaluate the
LSQ under uncertainty environments, so we pay little attention to the detailed indexes,
only considering five dimensions of SERVQUAL to evaluate. As a fact, the choice of
evaluation indexes is an interesting topic and we will study it in the future. In the
literature, the notions of IFSs and VSs are regarded as equivalent, in the sense that an IFS
is isomorphic to a VS (An and Wilfred, 2005). Although VSs and IFSs are equivalent by
basic definition, An and Wilfred (2005) show that VSs allow for a more intuitive
graphical representation of vague data, which facilitates significantly better analysis in
data relationships, incompleteness, and similarity measures. However, we do not compare
two methods in detail in the paper, thus an evaluation of LSQ by using IFSs will be
shown in future studies.

Acknowledgements

The authors would like to thank three anonymous reviewers for their helpful comments
and suggestions on earlier versions of this paper. This research work was supported in
part by the National Natural Science Foundation of China (NSFC) under the Grant
No.71203083 and 71033002.

References
An, L. and Wilfred, N. (2005) ‘Vague sets or intuitionistic fuzzy sets for handling vague data:
which one is better?’, in Delcambre, L., Kop, C., Mayr, H.C., Mylopoulos, J. and Pastor, O.
(Eds.): Conceptual Modeling – ER 2005: 24th International Conference on Conceptual
Modeling, Klagenfurt, Austria, 24–28 October, 2005, Lecture Notes in Computer Science,
Vol. 3716, pp.401–416.
Atanassov, K.T. (1986) ‘Intuitionistic fuzzy sets’, Fuzzy Sets and Systems, Vol. 20, No. 1,
pp.87–96.
Bellman, R.E. and Zadeh, L.A. (1970) ‘Decision-making in a fuzzy environment’, Management
Science, Vol. 17, No. 4, pp.141–164.
Bienstock, C.C., Mentzer, J.T. and Bird, M.M. (1997) ‘Measuring physical distribution service
quality’, Journal of the Academy of Marketing Science, Vol. 25, No. 1, pp.31–44.
Chandramathi, S., Raghuram, S.P.P., Srinivas, V.S. and Satyajit, S.H. (2008) ‘Dynamic bandwidth
allocation for 3G wireless systems: a fuzzy approach’, Applied Soft Computing, Vol. 8, No. 1,
pp.274–284.
Chang, C.W., Wu, C.R. and Lin, H.L. (2008) ‘Integrated fuzzy theory and hierarchy concepts to
evaluate software quality’, Software Quality Journal, Vol. 16, No. 2, pp.263–276.
Chen, S.H. (2000) ‘Representation, ranking, distance, and similarity of L-R type fuzzy number and
application’, Australian Journal of Intelligent Processing Systems, Vol. 6, No. 4, pp.217–229.
Chou, C.C. (2009) ‘A model for the evaluation of airport service quality’, Transport, Vol. 162,
No. 4, pp.207–213.
86 A-H. Wu et al.

Chou, C.C., Liu, L.J., Huang, S.F., Yih, J.M. and Han, T.C. (2011) ‘An evaluation of airline service
quality using the fuzzy weighted SERVQUAL method’, Applied Soft Computing, Vol. 11,
No. 2, pp.2117–2128.
Coyle, J.J., Bardi, E.J. and Langley Jr., J.C. (1992) The Management of Business Logistics, 5th ed.,
West, St. Paul, MN.
Gau, W.L. and Buehrer, D.J. (1993) ‘Vague sets’, IEEE Transactions on Systems Man and
Cybernetics, Vol. 23, No. 2, pp.610–614.
Grönroos, C. (1983) Strategic Management and Marketing in the Service Sector, pp.83–104,
Marketing Science Institute, Cambridge, MA.
Grönroos, C. (1984) ‘A service quality model and its marketing implications’, European Journal of
Marketing, Vol. 18, No. 4, pp.36–44.
Hsu, H.M. and Chen, C.T. (1997) ‘Fuzzy credibility relation method for multiple criteria
decision-making problems’, Information Sciences, Vol. 96, Nos. 1–2, pp.79–91.
Hu, Y.C. (2009) ‘Fuzzy multiple-criteria decision making in the determination of critical criteria
for assessing service quality of travel websites’, Expert Systems with Applications, Vol. 36,
No. 3, pp.6439–6445.
Irene, G.S. and Maria, E.R.M. (2011) ‘Logistics service quality and buyer-customer relationships:
the moderating role of technology in B2B and B2C contexts’, Service Industries Journal,
Vol. 31, No. 7, pp.1109–1123.
Kuo, M-S., Wu, J-W. and Pei, L. (2007) ‘A soft computing method for selecting evaluation criteria
of service quality’, Applied Mathematics and Computation, Vol. 189, No. 1, pp.241–254.
Liang, G.S. and Wang, M.J. (1991) ‘A fuzzy multiple criteria decision-making method for facilities
site selection’, International Journal of Production Research, Vol. 29, No. 11, pp.2313–2330.
Liang, G.S., Chou, T.Y. and Kan, S.F. (2006) ‘Applying fuzzy quality function deployment to
identify service management requirement for an ocean freight forwarder’, Total Quality
Management & Business Excellence, Vol. 17, No. 5, pp.539–554.
Mentzer, J.T., Flint, D.J. and Hult, T.M. (2001) ‘Logistics service quality as a segment-customized
process’, Journal of Marketing, Vol. 65, No. 4, pp.82–104.
Mentzer, J.T., Gomes, R. and Krapfel Jr., R.E. (1989) ‘Physical distribution service: a fundamental
marketing concept?’, Journal of the Academy of Marketing Science, Vol. 17, No. 1, pp.53–62.
Mikhailov, L. and Tsvetinov, P. (2004) ‘Evaluation of service using a fuzzy analytic hierarchy
process’, Applied Soft Computing, Vol. 5, No. 1, pp.23–33.
Montana, D. and Hussain, T. (2004) ‘Adaptive reconfiguration of data networks using genetic
algorithms’, Applied Soft Computing, Vol. 4 , No. 4, pp.433–444.
Ozcan, K. and Suzan, A.O. (2011) ‘Fuzzy AHP approach for supplier selection in a washing
machine company’, Expert Systems with Applications, Vol. 38, No. 8, pp.9656–9664.
Parasuraman, A., Zeithaml, V. and Berry, L. (1988) ‘SERVQUAL: a multi-item scale measuring
consumer perceptions of service quality’, Journal of Retailing, Vol. 64, No. 1, pp.12–40.
Perreault Jr., W.D. and Russ, F.A. (1974) ‘Physical distribution service: a neglected aspect of
marketing management’, MSU Business Topics, No. 3, pp.37–45.
Perreault Jr., W.D. and Russ, F.A. (1976) ‘Physical distribution service in industrial purchase
decisions’, Journal of Marketing, Vol. 40, No. 2, pp.3–10.
Shapiro, R.D. and Heskett, J.L. (1985) Logistics Strategy: Cases and Concepts, West, St. Paul,
MN.
Shih, Y.Y. and Hu, J.S. (2008) ‘Fuzzy quality attributes for evaluating internet marketing
system performance’, Total Quality Management & Business Excellence, Vol. 19, No. 12,
pp.1219–1234.
Stock, J.R. and Lambert, D.M. (1987) Strategic Logistics Management, 2nd ed., Irwin, Homewood,
IL.
A comprehensive evaluation of the logistics service quality 87

Sultan, F. and Simpson, M.C. (2000) ‘International service variants: airline passenger expectations
and perceptions of service quality’, Journal of Services Marketing, Vol. 14, No. 3,
pp.188–216.
Tsai, M.T., Wu, H.L. and Liang, W.K. (2008) ‘Fuzzy decision making for market positioning and
developing strategy for improving service quality in department stores’, Quality & Quantity,
Vol. 42, No. 3, pp.303–319.
Xu, J. (2008) ‘Logistics service quality analysis based on gray correlation method’, International
Journal of Business and Management, Vol. 3, No. 1, pp.58–61.
Yang, X.L., Zhou, J.ZH., Ding, J.H, Zou, Q. and Zhang, Y.CH. (2012) ‘A fuzzy AHP-TFN based
evaluation model of flood risk analysis’, Journal of Computational Information Systems,
Vol. 22, No. 8, pp.9281–9289.
Zadeh, L.A. (1965) ‘Fuzzy sets’, Information and Control, Vol. 8, No. 3, pp.338–353.
Zeithaml, V.A. (1988) ‘Consumer perceptions of price, quality, and value: a means-end model and
synthesis of evidence’, Journal of Marketing, Vol. 52, No. 3, pp.2–22.
Zhang, D.F., Huang, S.L. and Li, F. (2004) ‘An approach to measuring the similarity between
vague sets’, Journal of Huazhong University of Science and Technology, Vol. 32, No. 5,
pp.59–60.
Zhang, D.F., Zhang, J.L., Lai, K.K. and Lu, Y.B. (2009) ‘An novel approach to supplier
selection based on vague sets group decision’, Expert Systems with Applications, Vol. 36,
No. 5, pp.9557–9563.
Zimmermann, H.J. (1991) Fuzzy Set Theory and its Applications, 2nd ed., Kluwer Academic
Publishers, Boston.

You might also like