Journal of Marine Science and Technology, Vol. 25, No. 3, pp.

299-310 (2017 ) 299

DOI: 10.6119/JMST-016-1227-1

KEY INFLUENCE FACTORS FOR

OCEAN FREIGHT FORWARDERS SELECTING
CONTAINER SHIPPING LINES USING THE
REVISED DEMATEL APPROACH

Tien-Chun Ho1, Rong-Her Chiu2, Cheng-Chi Chung3, and Hsuan-Shih Lee4

Key words: sea transport, maritime marketing, container shipping DEMATEL could improve the shortcomings of DEMATEL. In
lines (CSLs), ocean freight forwarders (OFFs). conclusion, 12 key factors are proposed, and ‘integrated logistics’
and ‘timely delivery’ are the main influencing and consequence
factors respectively.
ABSTRACT
More than 80% of container traffic in the global container
I. INTRODUCTION
shipping market in recent years is derived from ocean freight
forwarders (OFFs). It raises concern about the OFFs’ role in co- Since the development of container transport in the 1960s,
ordinating the services between container shipping lines (CSLs) the environment of global container transport business has be-
and shippers. Most previous studies did not distinguish the iden- come increasingly competitive because of economic globaliza-
tity of shippers among OFFs, direct shippers, and routing order ex- tion and various trends in trade. Market uncertainty and unpre-
porters. There have been some studies examining the factors for dictable shipping changes have thus become the main factors
OFFs selecting CSLs by the method of decision-making trial and behind the impact of the choice of ocean freight forwarders
evaluation laboratory analysis (DEMATEL), but the initial direct (OFFs) on container shipping lines (CSLs). However, after 1995
relationship matrix may not have convergence to zero in the due to technological advances in shipbuilding, there was excess
original version of DEMATEL. Moreover, none of the previous space and supply shortage due to the development of mega-ships
studies applied the Revised DEMATEL analysis to evaluate the for global container shipping. Since then, the environment of glo-
key influence factors for OFFs selecting CSLs. This paper im- bal container transport business has become highly competitive
plemented a questionnaire survey of 30 experts from 15 major due to the financial crises of 2008 and 2012, and with the slow-
Taiwanese OFFs. The survey considered marketing 4C frame- ing of China’s economic growth since 2015; CSLs are facing
works of customer needs, customer costs, customer communica- new challenges in the current global economy. To counteract this,
tion, and customer convenience. The influence factors of maritime through the concept of market segmentation, CSLs should bet-
service for CSLs were constructed; the Modified Delphi Method ter understand the needs of shippers in order to enhance their
(MDM) and the Revised DEMATEL were used to define the satisfaction and operational performance (Wen and Lin, 2016).
suitability of key factors and to compare the different relations In practice, shippers can be divided into OFFs, direct ship-
among factors for OFFs in selecting CSLs to provide shipping pers, and routing order exporters. In the related research on choice
services. In particular, the result of research on the key factors of CSLs for Taiwanese shippers, most previous studies found
of the selection of CSLs by OFFs confirms that the Revised that shippers were unable to distinguish their identity (Krapfel
and Mentzer, 1982; Brooks, 1985; Brooks, 1990, 1991; Murphy
and Hall, 1995; Tiwari et al., 2003; Yen and Chen, 2004; Douglas
Paper submitted 03/21/16; revised 08/16/16; accepted 12/27/16. Author for et al., 2006; Salleh, 2007; Zsidisin et al., 2007; Brooks and Trifts,
correspondence: Cheng-Chi Chung (e-mail: [email protected]).
1
Ph.D. Student, Dept. of Shipping and Transportation Management, National
2008; Rogerson et al., 2014). Chen et al. (2009) showed that
Taiwan Ocean University, Keelung, Taiwan, R.O.C. more than 80% of the container traffic in the global container
2
Associate Professor, Dept. of Shipping and Transportation Management, shipping market was from OFFs, and it has aroused concern
3
National Taiwan Ocean University, Keelung, Taiwan, R.O.C. about who being in charge of the service between CSLs and
Professor, Dept. of Shipping and Transportation Management, and Chairman, shippers. In addition, OFFs’ subjective response in the quality
Undergraduate Program of Ocean Tourism Management, National Taiwan
Ocean University, Keelung, Taiwan, R.O.C.
services of CSLs is less effective than that of direct shippers.
4
Professor, Dept. of Shipping and Transportation Management, National Taiwan The current study points out those better shipment of goods,
Ocean University, Keelung, Taiwan, R.O.C. accompanying more satisfaction of OFFs, could strengthen the
300 Journal of Marine Science and Technology, Vol. 25, No. 3 (2017 )

partnership between CSLs and OFFs. Therefore, it leads to bet- provides a comprehensive discussion.
ter applications in shipping business. In principle, freight tariffs
are the most important consideration for OFFs in choosing a 1. Attributes of Maritime Service for Container
carrier (McGinnis, 1990). In addition to considering freight tariff, Shipping Lines
OFFs also consider service quality of sailing accuracy (Yang Collison (1984) considered delivery speed, reliable schedules,
et al., 2014), shipping security, cargo tracing system and trans- freight tariff, and cargo damage claims as the primary service
feral of transports (Wong et al., 2008; Chung et al., 2011). As factors for studying the choice of OFFs for CSLs. The freight
CSLs are mostly based on OFF’s orientation, they should under- charges did not vary greatly between different carriers during
stand the shipping market changes and shippers’ needs. If CSLs the 1980s, carrier reputation (Brooks, 1985) and customs clear-
can better understand the dynamic changes of customer needs, ance efficiency (Slack, 1985) were the main determining fac-
as well as ways to reduce customer cost, improve customer com- tors for the choice of OFFs for CSLs at that time. Subsequently
munications and provide more convenient services, they can im- Brooks (1990, 1991) showed that the importance of the transit
prove the operation performance in the uncertain shipping market. time was greater than carrier reputation in the early 1990s. Also
The Delphi technique is a method for establishing a group in that decade, companies’ financial key performance indicator
communication process, allowing a group of individuals, as a (KPI) reports, together with an increasing trend to use branch
whole, to deal with a complex problem. This structured com- offices and commission agents, generated an expanded scope of
munication is accomplished through various feedbacks, includ- services, and providing higher quality service became an impor-
ing individual contributions of information and knowledge. It tant feature for OFFs (McGinnis, 1990).
consists of assessment of the group judgments or views, op- Freight tariffs have been the highest priority consideration
portunities for individuals to revise views, and a degree of ano- since 2000, because CSLs had greater room for bargaining over
nymity for the individual responses (Linston and Turoff, 1975). price, this will increase an OFF’s intention to use that CSL for
The modified Delphi Method (MDM) retains the spirit and shipment (Shry and Chu, 2005). With the vigorous developments
advantages of original Delphi method, and also makes two in the logistics industry, CSLs needed to integrate their services
further improvements. First, it develops a structured question- vertically, and increase the efficiency of transport processes
naire to replace the complicated survey used in the traditional (Douglas et al., 2006). It has been noted that CSLs can enhance
Delphi open-ended questionnaire, allowing the experts to focus their market competitiveness by integrating their logistics opera-
on the research topics and improve the overall response rate. tions (Tiwari et al., 2003; Huang, 2014; Yang et al., 2014).
Secondly, it uses statistical analysis and systematic data pro- Furthermore, considering the need of OFFs for logistics op-
cessing to integrate the experts’ opinions and reach a consensus erations, the efficiency of pier operations and fees are also key
of these views. Subsequently, the decision-making trial and eva- factors considered by shippers (Tongzon, 2009).
luation laboratory (DEMATEL) has been applied in many fields, Under the pressure of intensive global competition, supported
such as marketing strategies, control systems, safety problems, by the use of e-commerce, the container shipping lines can pro-
developing the competencies of global managers and group de- vide benefits for customers with more value-added services
cision making. However, this method, raising the initial relation (Penaloza et al., 2007). The increased effectiveness from pro-
matrix to the power of infinity, may not yield a convergence to viding e-tracking systems, rapid response, reduction of cargo da-
zero and hence total influence may not converge, though the mage, and enhancement of transport safety (Liang et al., 2007)
Revised DEMATEL can improve on this shortage (Lee et al., can increase the loyalty of their shippers. According to Wong
2013). et al. (2008), voyage reliability and communication skills are
As mentioned above, the current study applies more strin- more important than freight tariffs because transport delay and
gent subcriteria for the selection process by using the MDM inefficient staff are the most frequently encountered problems.
and Revised DEMATEL, not only to define the suitability and Hence, service attitude and maintaining relationships are also
relevance of key factors and affecting factors, but also to adopt key factors for the choice of OFFs by CSLs.
the 4Cs criteria of customer orientation to compare OFFs se- As transport providers are part of the service industry, it is
lections of CSLs. The rest of this paper is organized as follows. necessary for CSLs to understand the different needs of OFFs
Section 2 reviews the related literature on the choice of OFFs in order to provide satisfactory service. In addition to reliable
for CSLs and their proposed consolidation. Section 3 explores transport, CSLs should consider transport safety, lower costs, ex-
methods and the assessment framework. Section 4 provided pertise, company reputation, transit time (Brooks and Trifts,
an empirical analysis of OFFs’ choices for CSLs. Finally, con- 2008), service scope, integrated logistics (Krapfel and Mentzer,
clusions and recommendations are presented in section 5. 1982; Yeung, 2006), intensive sailing frequencies (Notteboom,
2006), and implementation of e-commerce (Wen and Lin, 2016).
In addition, other aspects such as container types, sizes, conven-
II. LITERATURE REVIEW
ient capacity and accurate documentation are also key influencing
This section reviews attributes related to maritime service factors for OFFs to choose CSLs (Yen and Chen, 2004; Chung
and discusses the impact of various factors on choosing CSLs. et al., 2011).
It also reviews the related literature on research methods and In the research on how shippers choose CSLs, most previous
 T.-C. Ho et al.: Key Factors for OFFs Selecting CSLs Using RDEMATEL 301

studies were unable to distinguish the types of shippers, i.e., all factors, thereby clarifying the complex causal relationship
OFFs, direct shippers, and routing order exporters, for evalu- between the evaluation criteria and the decision-making goal.
ation criteria. Wen and Huang (2007) and Wen and Lin (2016) It has been used to solve sophisticated problems by improving
studied on OFFs to selecting CSLs, but there was a lack of the understanding of them (Tzeng et al., 2007). Using mutual
theoretical support for the criteria. Kannan (2010) and Kannan relations between the factors of comparison to calculate the
et al. (2011) used SERVQUAL framework, factor analysis (FA) direct, indirect, and combined effects, this method helps clar-
and an analytic hierarchy process (AHP) approach to analyze ify the nature of problems and resolve related issues (Liu and
ocean carriers’ strengths and weaknesses in India, though they Lin, 2005). Since the initial direct relationship matrix may not
did not analyze the relevance of subcriteria. Chung et al. (2011) have convergence to zero in the original version of DEMATEL,
used 7Ps and DEMATEL to assess the key factors of Taiwanese the Revised DEMATEL improves on this shortage (Lee et al.,
OFFs selection of CSLs. However, the initial relation matrix, 2013). Yang (2013) used the Revised DEMATEL to analyze di-
raised to the power of infinity, may not converge to zero. More- rectly and indirectly, and combined factors for choosing a ship’s
over, since the total influence may not converge, the results registration, which provides an effective strategy for CSLs.
will probably be unable to effectively present the correlations In place of DEMATEL, this study adopts Revised DEMATEL,
between subcriteria. Briefly, previous studies have not been supplemented by MDM to survey the optimum subcriteria ef-
able to comprehensively consider OFFsʼ and the theoretical basis fectively and objectively. This study focuses on senior manager’s
of criteria is lacking. Since it is difficult to comprehensively opinion from 15 major Taiwanese OFFs to select appropriate
analyze the current situation of the container shipping market, evaluation subcriteria, to establish the influence factors for OFFs
this leads to ineffectively applying value on the basis of eva- selecting CSLs, and to provide shipping companies practical
luation criteria and shipping management that are incompletely value as well as a reference for future research. The results of
understood. In order to compensate for shortcomings in pre- research into the key factors of the selection of CSLs by OFFs
vious studies, this study includes customer needs, customer confirm that the Revised DEMATEL can improve the short-
costs, customer communication and customer convenience as comings of DEMATEL.
the influence factors for analyzing OFFs selection of CSLs.
This provides a more complete understanding of the relevance 3. Comprehensive Discussions
of influence factors for OFFs selecting CSLs in Taiwan. Although previous studies have considered the influencing
factors for OFFs to choose CSLs, their needs will vary accord-
2. Related Literature and Research Methods ing to conditions, and CSLs need to understand this in order to
The Delphi method, proposed by Dalkey and Helmer in maintain existing customers and develop new ones (Chung
1960, is a systematic method to expedite decisions of expert et al., 2011). Relevant evaluation of subcriteria and descriptions
group. Murry and Hammons (1995) suggested that the Delphi of influencing factors in maritime services for CSLs are shown
method should adopt an anonymous decision-making technique in Table 1.
to obtain input from a group of experts. Although this can effec- In conclusion, for the optimal transport conditions, CSLs should
tively collect opinions, but it is difficult to control research understand how attributes of OFFs differentiate the needs and
progress and there is a low rate of survey returns. The MDM affect the strategies for shipping services. Moreover, because
was then proposed to resolve these problems. In the MDM, re- DEMATEL models the influences of components of a system
levant literature can be considered to modify the speculation with an initial direct relation matrix, so the influences of com-
that occurs in traditional open-ended Delphi questionnaires and ponents can transitively affect other components; and this is
allow experts to focus more clearly on the research topics. The modeled by raising the initial direct relation matrix to powers.
MDM has been extensively applied, based on the expert group’s The total influence is computed by summing up the matrices
work experiences and knowledge as expressed in the question- of all powers based on the assumption that the matrix raised to
naires. For practical applications, Lirn et al. (2004) proposed the power of infinity would converge to zero. Lee et al. (2013)
developing evaluation criteria for influence transshipment fac- indicated that if the initial relation matrix, raised to the power
tors using the MDM. Hsu (2010) applied the MDM to help of infinity, does not converge to zero, then the total influence
investors effectively select an optimal location for an interna- may not converge.
tional business center in China, a case that is highly relevant for This study is based on customer orientation, considering the
both academic and commercial implications. Lin et al. (2011) relevant literature on how OFFs choose CSLs, and this paper
explored the best option for the location of an international ex- effectively improves the evaluation subcriteria for CSLs. MDM
hibition, and established evaluation criteria and subcriteria. is used to develop stringent selection subcriteria and to gain valu-
Subsequently, the Battelle Memorial Institute of the Geneva Re- able and objective data information through expert groups’ re-
search Centre between 1972 and 1976 developed the DEMATEL levant work experiences and knowledge. Based on this, the Re-
approach. An initial direct relation matrix was established to vised DEMATEL ensures that the initial direct-relation matrix
observe the degree of interaction between factors. By applying to infinite power will converge to zero, and thereby defines the
it, the matrix and related mathematical theory can be used to suitability and relevance of key factors for CSLs to create op-
calculate the causal relationship between influence degrees of erating strategies.
302 Journal of Marine Science and Technology, Vol. 25, No. 3 (2017 )

Table 1. Relevant subcriteria of influence factors in maritime service for CSLs.

Influencing factors Definitions Sources
Used to measure the capacity of CSLs to make a profit in a par-
Financial KPI report ticular period, such as concerning yielding rate, return on assets, McGinnis (1990)
total assets turnover, return on equity, current ratio, debt ratio, etc.
CSLs advertise transport services through the intermediate trader,
Indirect access network Chung et al. (2011); Yeung (2006)
such as OFFs or brokers
Participate in port investment and harbor leasing to meet CSLs’
Yang et al. (2014); Chung et al. (2011); Tongzon (2009);
Dedicated dock needs, pre-scheduling to improve efficiency in using dock machin-
Vernimmen et al. (2007); Notteboom (2006)
ery and equipment and to ensure shipment stability
Processing costs for receiving, amending and surrendering cer-
Document fee Chung et al. (2011)
tificates of origin and other documents
Providing different sizes and functions of container to meet ship- Wen and Lin (2016); Chung et al. (2011); Wong et al.
Types and condition
pers’ various needs. Containers structure should conform to be (2008); Wen and Huang (2007); Yen and Chen (2004);
of Container
standard and internally clean to ensure transport safety Brooks (1995)
High-speed and automated mega container ships, reduce the unit
Yang et al. (2014); Chung et al. (2011); Mentzer et al.
Mega container ships cost of container transport and increase the convenience of ship-
(1999)
pers to utilize space
CSLs should seek to document accuracy to shorten the time taken Wen and Lin (2016); Chung et al. (2011); Wen and
Document accuracy
between CSLs and shippers Huang (2007); Yen and Chen (2004); Brooks (1995)
Wen and Lin (2016); Wong et al. (2008); Liang et al.
Equipment obtained Including the withdrawal of general and special containers, as well
(2007); Wen and Huang (2007); Shry and Chu (2005);
conveniently as conveniently obtaining spaces
Yen and Chen (2004); Brooks (1995)
Wen and Lin (2016); Chung et al. (2011); Kannan et al.
Shippers can use online booking and tracking to monitor the flow
E-commerce system (2011); Kannan (2010); Wong et al. (2008); Penaloza
of goods via the internet
et al. (2007); Yen and Chen (2004)
Chung et al. (2011); Kannan et al. (2011); Kannan
Service attitude Timely solution of problems with patience and a helpful attitude
(2010); Wong et al. (2008); Wen and Huang (2007)
Booking, shipping, issuance documents and withdrawal of cargo Yang et al. (2014); Chung et al. (2011); Liang et al.
Convenient shipping
are simple and convenient (2007); Yen and Chen (2004)
Wen and Lin (2016); Yang et al. (2014); Chung et al.
Regularly visit responsible shippers, establish business relation-
Relationships maintaining (2011); Wen and Huang (2007); Yen and Chen (2004);
ships, solve problem during the implementation
Tiwari et al. (2003)
CSLs set up branches abroad in order to facilitate direct shipper
Direct access network Chung et al. (2011); Yen and Chen (2004)
inquiry for transport services
Intensive use of print media, including published shipment ad-
Sailing schedule advertisement Chung et al. (2011)
vertising transport services to targeted customers
Wen and Lin (2016); Rogerson et al. (2014); Chung
et al. (2011); Kannan et al. (2011); Kannan (2010); Wong
et al. (2008); Liang et al. (2007); Salleh (2007); Wen
Freight tariffs Shippers pay the freight costs to CSLs for ocean transport and Huang (2007); Douglas et al. (2006); Shry and Chu
(2005); Mentzer et al. (1999); Brooks (1995); Murphy
and Hall (1995); Brooks (1990); Brooks (1985); Krapfel
and Mentzer (1982)
Due to increased transport services, excess delivery time will cause Wong et al. (2008); Liang et al. (2007); Penaloza et al.
an excessive increase of downstream industry demand conditions, (2007); Salleh (2007); Vernimmen et al. (2007); Douglas
Timely delivery
and will also affect the willingness of importers to order from ex- et al. (2006); Notteboom (2006); Shry and Chu (2005);
porters Liao et al. (2004)
Wen and Lin (2016); Chung et al. (2011); Kannan et al.
(2011); Kannan (2010); Brooks and Trifts (2008); Salleh
The reliability and accuracy of time and shipment for transport
(2007); Wen and Huang (2007); Vernimmen et al. (2007);
Transport reliability services. Scheduled time, timetable stability, arrival and estimated
Notteboom (2006); Liao et al. (2004); Yen and Chen
arrival time are consistent
(2004); Mentzer et al. (1999); Brooks (1995); Brooks
(1991); Brooks (1985); Collison (1984)
 T.-C. Ho et al.: Key Factors for OFFs Selecting CSLs Using RDEMATEL 303

Table 1. Relevent subcriteria of influence factors in maritime service for CSLs (cont.).
Influencing factors Definitions Sources
Wen and Lin (2016); Yang et al. (2014); Chung et al.
(2011); Tongzon (2009); Brooks and Trifts (2008); Wong
The amount of vessels, ports of call frequency and the number of
Sailing frequency et al. (2008); Vernimmen et al. (2007); Wen and Huang
departures
(2007); Notteboom (2006); Brooks (1995); Brooks (1990);
Slack (1985); Brooks (1985); Collison (1984)
Huang (2014); Rogerson et al. (2014); Yang et al. (2014);
An logistics company combines inland and maritime services for
Chung et al. (2011); Vernimmen et al. (2007); Yeung
Integrated logistics intermodal transport. Development and integrated international lo-
(2006); Yen and Chen (2004); Tiwari et al. (2003); Heaver
gistics operation mode, providing consistent door-to-door services
(2001); Brooks (1985); Krapfel and Mentzer (1982)
Huang (2014); Yang et al. (2014); Brooks and Trifts
Transit time Refers to time spent for goods in transport (2008); Wong et al. (2008); Douglas et al. (2006); Brooks
(1990); Slack (1985); Brooks (1985); Collison (1984)
Providing professional advice on shipper transport, and capacity Wen and Lin (2016); Chung et al. (2011); Wong et al.
Maritime expertise of transport-related problem processing with transport and logistics (2008); Liang et al. (2007); Wen and Huang (2007);
expertise Yen and Chen (2004); Brooks (1995); Brooks (1985)
Wong et al. (2008); Liang et al. (2007); Shry and Chu
CSLs shall conform to laws and regulations of the contract and
Transport security (2005); Yen and Chen (2004); Slack (1985); Brooks
pay more attention to ensure the ship and cargo safety
(1985)
Rogerson et al. (2014); Chung et al. (2011); Vernimmen
Combination of Most appropriate route combinations are based on the arrange-
et al. (2007); Douglas et al. (2006); Liao et al. (2004);
operating route ment and efficiency of direct or transshipment services
Brooks (1985)
In order to establish a good image and reputation, CSLs should
Wen and Lin (2016); Chung et al. (2011); Kannan et al.
actively participate in social public benefit activities, implemen-
Reputation and image (2011); Kannan (2010); Wen and Huang (2007); Yen
tations of green energy and environmental protections, as well as
and Chen (2004); Brooks (1991); Slack (1985)
fulfilling corporate social responsibilities.
Recurrent or non-recurrent charges, such as bunker adjustment
Rogerson et al. (2014); Yang et al. (2014); Chung et al.
Surcharges factor, currency adjustment factor, peak season surcharge, port con-
(2011); Tongzon (2009); Slack (1985)
gestion surcharge and inland transport costs
Good overall image of CSL contacts with the shipper to develops
Staff appearance a positive impression by the shipper, e.g., being well groomed and Wen and Lin (2016); Chung et al. (2011); Brooks (1985)
appropriately dressed
Recurrent charges or additional surcharges for loading and un-
Loading and unloading charges loading in port, including terminal handling charges, container Chung et al. (2011); Slack (1985)
freight station charges, as well as demurrage and detention
The proportion of goods to be inspected by customs, together with
Customs clearance efficiency clearance efficiency, relationship with customs and cargo with follow- Chung et al. (2011); Slack (1985)
up treatment of items detained by customs

III. RESEARCH METHODS AND opinions without interference. With statistical analysis and sys-
EVALUATION FRAMEWORK tematic processing methods, an archive can facilitate the expert
group to develop its evaluation subcriteria of influence factors
for maritime service of CSLs. The collection of expert opinions,
The following is a description of research methods used
determination of consistency, stability and consent determi-
in this paper, together with the influence factors of evaluation
nation was as follows (Lee et al., 2008).
framework.
1) Collection of Expert Opinions
1. Modified Delphi Method According to the methods mentioned above, this study used
There are numerous related service impact factors affecting three rounds of questionnaires to collect experts’ opinions, and
CSLs. To explore the most appropriate evaluation subcriteria thereby understand the factors influencing OFFs’ selection of
and reach an agreement between experts, this study adopts the CSLs. In order to understand the extent to which the experts
Modified Delphi Method (MDM) to develop the structured ques- agree on the description of each topic, a Likert scale was used
tionnaire. A group of experts synthesizes the senior managers’ to evaluate reactions of the experts’ opinions, supplemented by
304 Journal of Marine Science and Technology, Vol. 25, No. 3 (2017 )

collecting individual experts’ opinion. For questionnaire statistics, 3) Calculate the Normalized Initial-Direct Relation Matrix X,
this research used the quartile deviation test on the variation of Which Is Calculated by
individual observations to understand the consensus of all experts.
A
2) Determination of Consistency X  (2)
s
Faherty (1979) indicated that a quartile deviation of less than
or equal to 0.6 can be considered to indicate that the opinions
where
of experts reached a high level of consistency. The quartile de-
viation from 0.6 to 1.0 indicates that the experts’ opinions reached n n
a moderate degree of consensus for this topic. And if a quartile s  max(max  aij ,   max  aij ) (3)
1 i  n 1 j  n
deviation, greater than 1.0 indicates the topic does not reach a j 1 i 1
consensus. On the consistency test, if more than 85% topics
reached a high or moderate level of consensus, the questionnaire and  is a very small positive number.
could be considered as completed.
4) Derive the Total Influence Matrix S
3) Determination of Stability
All indirect influence matrices are X2, X3, , Xk, , X the
When a topic does not reach consensus, Murry and Hommons total influence matrix, which is equal to
(1995) suggested that there is small possibility, lower than 20%,
of experts altering their opinions. This situation showed the con-
sistency and stability. S  X ( I  X ) 1 (4)

4) Consent Determination 3. Evaluation Framework of Influence Factors

“Strongly disagree”, “disagree”, “neither agree nor disagree”, Moeller and Shafer (1987) proposed that the experts should
“agree” and “strongly agree” are the format of a typical 5-point be selected based on their experience, knowledge, reputation
Likert item in this study. Regarding how most of experts make and willingness to cooperate. For achieving satisfaction in eva-
appropriate judgement, we applied statistical mode to represent luation subcriteria, OFFs objectively gathered the influence fac-
expert opinions. tors for choosing CSLs, based on the reviewed literature. The
2. The Revised DEMATEL combined marketing 4C frameworks proposed by Lauterborn
(1990) include structures for customer needs, customer costs,
The Battelle Memorial Institute of the Geneva Research Centre
customer communication and customer convenience. Regarding
between 1972 and 1976 developed the DEMATEL approach.
the scope of evaluations, criteria and subcriteria from this study,
The approach establishes an initial direct relation matrix to ob-
Fig. 1 shows abstract influence factors for maritime service from
serve the degree of interaction between factors, and the matrix
CSLs.
and related mathematical theory are used to calculate the causal
relationship between influence degrees of all factors, thereby
clarifying the complex causal relationship between the evalu- IV. EMPIRICAL ANALYSIS AND DISCUSSIONS
ation criteria and decision-making goals (Seyed-Hosseini et al.,
2006; Hsu et al., 2013). Since the initial direct relationship matrix This section is based on questionnaires filled out by the ex-
may not converge to zero in the original version of DEMATEL, perienced and knowledgeable shipping industry experts. Based
this is improved by the Revised DEMATEL (Lee et al., 2013). on this data, we use Excel software to analyze the relevance and
It is calculated as follows. correlation of the influence factors for OFFs to choose CSLs.
1) Define and Determine the Relationship Between the Factors 1. Survey Results
Filter and define factors in the system according to experts’ The questionnaire survey of this study is divided into two
experience and literature review. stages, first by mailing surveys and then by direct questionnaires
2) Calculate the Initial Average Matrix and interviews. Murry and Hammons (1995) and Rowe and
Wright (1999) stated that the most appropriate number of ex-
Let A  (aij ) n  n be an average matrix of the respondents’ perts should be between 10 and 30 when applying the Delphi
direct matrices in which the entry (i, j) indicates the direct in- method. If there are more than 30, this will lead to complications
fluence that factor i exerts on factor j. The initial average matrix and a greater workload. It creates difficulties in obtaining valid
A  (aij ) n  n is given by conclusions. Accordingly, this study uses convenience sam-
H
pling focusing on 15 major OFFs. Participants in this survey,
1
A
H
B
k 1
(k )
(1) including general managers, deputy general managers, and man-
agers, are responsible for selecting CSLs for their corporations.
For the first, second and third stages 30, 26 and 21 question-
where B(k) is the answering matrix of the k-th respondent. naires were distributed, with effective questionnaire return rates
 T.-C. Ho et al.: Key Factors for OFFs Selecting CSLs Using RDEMATEL 305

Transport reliability
Combination of operating route
Dedicated dock

Shipper’s need Integrated logistics

Types and condition of container
Transport security
Freight tariffs
Loading/unloading charges
Surcharge
Document fees
Shipper’s costs
Transit time
Timely delivery
Key Documentation accuracy
influence
factors on Sailing schedule advertisement
maritime E-conmerce system
service for
container Reputation and image
shipping lines
Staff appearance
Shipper’s communication
Service attitude
Relationships maintaining
Maritime expertise
Financial KPI report
Direct access network
Indirect access network
Mega container ships
Shipper’s convenience Convenient shipping
Equipment obtained conveniently
Sailing frequency
Customs clearance efficiency

Goal Criteria Sub-criteria

Fig. 1. Evaluation framework of influence factors on maritime service for OFF’s selecting CSLs.

of 86.67%, 80.77% and 80.95%. The screening processes on reliability,’ ‘integrated logistics,’ ‘transport security,’ ‘freight
influence factors for OFFs selecting CSLs are shown in Table 2. tariffs,’ ‘transit time,’ ‘timely delivery,’ ‘service attitude,’ ‘mari-
time expertise,’ ‘direct access,’ ‘convenient shipping,’ ‘sailing
2. Overall Assurance Analysis of Influence Factors frequency,’ and ‘customs clearance efficiency.’ These are used to
Faherty (1979) indicated that a quartile deviation of less than process the correlation analysis of key influence factors in mari-
or equal to 0.6 can be considered that the opinions of experts time service for CSLs. 
reached a high level of consistency. A significance level equal
to or above 3.5 can be considered that the opinions of experts 3. Correlation Analysis of Influence Factors
can be accepted (Chen and Chen, 2011). The quartile deviation An example is illustrated in this section. Let us revisit the
in this study is less than or equal to 0.6 and the significance is example by the Revised DEMATAL, as follows. Lee et al. (2013)
above 3.5, indicating that the expert opinions reached consis- let B(1) and B(2) are the matrices of a system that are given by
tency, and so the evaluation subcriteria are retained. From the two experts. The answer matrices corresponding to the matrixes
evaluation subcriteria, this study obtained 12 factors: ‘transport are as follows:
306 Journal of Marine Science and Technology, Vol. 25, No. 3 (2017 )

Table 2. Screening process on influence factors for OFFs selecting CSLs.

First round Second round Third round
Evaluation factors
assurance consistency result assurance consistency result assurance consistency result
Transport reliability 4 0.5 retention - -
Combination of operating route 3.25 0.625 observed 3.75 0.5 delete -
Dedicated dock 3 1 observed 3 0.625 observed 3 0.5 delete
Integrated logistics 3 0.625 observed 4 0.5 retention -
Types and condition of container 3 0.625 observed 3 0.5 delete -
Transport security 4 0.5 retention - -
Freight tariffs 5 0 retention - -
Loading/unloading charges 3 0.5 delete - -
Surcharges 4 1 observed 3 0.625 observed 3 0.5 delete
Document fees 3.25 0.625 observed 3 0.5 delete -
Transit time 4 0.5 retention - -
Timely delivery 4 0.5 retention - -
Documentation accuracy 3 0.5 delete - -
Sailing schedule advertisement 2 1 observed 2 0.5 delete -
E-commerce system 3 0.5 delete - -
Reputation and image 3 0.5 delete - -
Staff appearance 3 0.5 delete - -
Service attitude 4 0.5 retention - -
Relationships maintaining 3 0.5 delete - -
Maritime expertise 4 0.5 retention - -
Financial KPI report 2.75 0.625 observed 4 0.625 observed 3 0.5 delete
Direct access network 4 0.5 retention - -
Indirect access network 3.25 0.125 delete - -
Mega container ships 3 1 observed 2.75 0.5 delete -
Convenient shipping 4 0.5 retention - -
Equipment obtained conveniently 4 0.625 observed 3 0.375 delete -
Sailing frequency 4 0.5 retention - -
Customs clearance efficiency 4 0.5 retention - -

0 4 2 0 0 3 0 1  0 0.6999986 0.1999996 0.0999998 

4 0 0 
1 4 0 1 0   
B (1)  and B (2)    0.7999984 0 0.0999998 0.0999998 
1 0 0 4 1 0 0 4 X  
     0.1999996 0 0 0.7999984 
0 2 4 0 0 1 3 0  
 0 0.2999994 0.6999986 0 
Step 1. Step 3.
Initial average influence matrix is Since

 0 3.5 1 0.5
 4 0 0.5 0.5 0 0 0 0
 
A 
0 0 0 0
1 0 0 4  
X 
  ,
 0 1.5 3.5 0  0 0 0 0
 
0 0 0 0
Step 2.
Let  = 10-5. initial influence matrix is we have
 T.-C. Ho et al.: Key Factors for OFFs Selecting CSLs Using RDEMATEL 307

Table 3. Correlation values of key influence factors.

Influence affecting factors Dk Rk Dk  Rk (ranking) Dk  Rk (ranking)
C11 transport reliability 1.094 1.406 2.500 (2) -0.321 (5)
C14 integrated logistics 1.673 0.994 2.667 (1) 0.679 (1)
C16 transport security 0.311 0 0.311 (7) 0.311 (3)
C21 freight tariffs 0 0.332 0.332 (6) -0.332 (6)
C25 transit time 1.010 0.635 1.645 (4) 0.375 (2)
C26 timely delivery 0.668 1.405 2.073 (3) -0.737 (7)
C47 customs clearance efficiency 0.320 0.303 0.623 (5) 0.017 (4)

D-R

0.8
0.7 C14
0.6
0.5
0.4
C16 C25
0.3
0.2
0.1
0 C47 D+R
-0.1 0.5 1.0 1.5 2.0 2.5 3.0
-0.2
-0.3 C21 C11
-0.4
-0.5
-0.6
-0.7
C26
-0.8

Fig. 2. Cause and effect relationship of key influence factors.

S  X  X 1  X 2  X   the factor is an influence factor, whereas a negative number

means the factor is influenced. A figure indicating the cause and
125000.0321 125000.397 124999.8209 124999.7497  effect relationship in the key influence factors given a set thresh-
125000.5204 125000.026 129999.7497 124999.7043  old is presented in Table 3 and Fig. 2.
 Table 3 shows that after a given threshold, the high degree
124999.7303 124999.717 125000.0321 125000.5204 
  of correlation factors include ‘C14 integrated logistics,’ ‘C11
124999.7173 124999.860 125000.3973 125000.0256  transport reliability,’ ‘C26 timely delivery,’ ‘C25 transit time,’
‘C47 customs clearance efficiency,’ ‘C21 freight tariffs,’ and
For the correlation of key influence factors, in order to ob- ‘C16 transport security’.
tain stronger influence factors, 0.14 is used as the threshold in this
study. This is completed to remove any factors that have a low cor- 4. Managerial Implications
relation since a direct or indirect relationship value greater than Compared with other main influence factors, ‘integrated lo-
0.14 indicates a greater significance effect. Therefore, the sum of gistics’ is the primary factor, and it can be bidirectionally affected
determinant by each row and column calculates the total extent by ‘transit time,’ ‘timely delivery’ and ‘transport reliability’.
of affected and ranking in the key influence factors. Di indicates Followed by ‘transit time,’ it can be bidirectional affected by
the extent of the factor’s influence on other factors, Rj indicates ‘transport reliability’ and ‘integrated logistics’. Due to the rise
the extent of the factor receiving influence, and (Dk  Rk) in- of logistics and transport services, logistics providers are part
dicates the intensity of the factor with others. A larger value in- of the transport industry. Along with the increase in the chang-
dicates that the total impact factor is greater. (Dk  Rk) indicates ing of transport conveyors, it will also raise the transit time and
the extent of factors’ interaction. A positive number indicates that probability of damaged goods, so CSLs will be involved with
308 Journal of Marine Science and Technology, Vol. 25, No. 3 (2017 )

different transport services and units for providing consistent also provide ‘integrated logistics’ and ‘transport reliability’ in
logistics services. If a CSL can integrate logistics effectively, order to attract OFF’s shipment attention.
reduce transit time, enhance ‘timely delivery’ and ‘transport re-
liability,’ reduce complex shipping activities handling by OFFs, V. CONCLUDING REMARKS
economize time, labor and expenditure, this will increase an
OFF’s intention to use that CSL for shipment. The global container transport business has become increas-
Yeung (2006) indicated that if CSLs attempt to gain market ingly competitive due to the increasingly global economy and
advantage, in addition to providing products and services, they greater trade demands. Market uncertainty and the shipping en-
should also have a complete logistics system. For CSLs, Zacharia vironment’s unpredictable changes may become important guide-
and Mentzer (2004) indicated that logistics are significantly ef- lines for shipper’s choice of CSLs (Murphy and Hall, 1995). This
fective in assisting enterprises, creating competitive advantage, study is based on the OFF’s orientation to establish an frame-
improving profitability and customer satisfaction. This provides work for the evaluation of influence factors in maritime service
more competitive ‘freight tariffs’ and shortens ‘transit time’ in for OFFs who are selecting CSLs. Based on shippers’ needs, ship-
order to attract OFF’s intention. Furthermore, ‘transport re- pers’ costs, shippers’ communication, and shippers’ convenience
liability’ can be unidirectional affected by ‘transport security’. as evaluation factors and 28 evaluation subcriteria, a question-
Due to the dramatic increase in global economic development naire survey for OFFs selecting Taiwanese CSLs was prepared
and trade capacity, with the growth of vessel quantity and the and administered. MDM was used to define the adaptation of
trend of maximizing vessel size, the demands on maritime ship- evaluation subcriteria and the Revised DEMATEL was used to
ping have not been reduced, even with improvements in marine define the relevance and suitability of key influence factors in
science and technology. Furthermore, CSLs can ensure the safety maritime service for CSLs.
of navigation and decrease damage to cargo, which will also create The analysis of influence factors affecting maritime service
their credibility to OFFs. for OFFs selecting CSLs included ‘transport reliability,’ ‘inte-
Subsequently, ‘transit time’ can be bidirectionally affected by grated logistics,’ ‘transport security,’ ‘freight tariffs,’ ‘transit time,’
‘transport reliability’. Because maritime transport takes more ‘timely delivery,’ ‘service attitude,’ ‘maritime expertise,’ ‘direct
time than other modes of transport, CSLs should conform the access,’ ‘convenient shipping,’ ‘sailing frequency,’ and ‘customs
provisions of international laws and safety regulations to main- clearance efficiency’. These 12 factors are the most appropriate
tain seaworthiness, schedule reliability and be consistent to actual evaluation subcriteria. Correlation analysis of influence factors
time of arrival and estimated time of arrival. Thereby it would for OFFs selecting CSLs, and the key influence factors are ‘in-
increase OFFs acceptance of the transit time and timely de- tegrated logistics,’ ‘transit time,’ ‘transport security’ and ‘customs
livery by CSLs. In other words, CSLs should not only care about clearance efficiency’. The key consequence factors are ‘timely
‘transport reliability’ or reducing ‘freight tariffs,’ if they can sat- delivery,’ ‘freight tariffs,’ and ‘transport reliability’. In the lite-
isfy both of those and other factors to enhance ‘timely delivery,’ rature, only Chung et al. (2011) analyzed the relevance of key
this will increase the likelihood of becoming priority for OFFs. influence factors for OFFs selecting CSLs, and the results in-
Among the consequence factors, ‘timely delivery’ is the most dicated that sales expertise could affect transport reliability. In
important, followed by ‘freight tariffs’ and ‘transport reliability’. contrast, the current study finds that transport reliability can be
Concerning ‘timely delivery,’ long transit time and shipment affected by transport security. It is shown that the OFFs might
delay will affect the subsequent customs clearance efficiency and be influenced by shipping accidents, such as the MOL and TSL
logistics operations, which in turn will affect the overall ship- in Hong Kong and Taiwan. Considering transport fees, this study
ment scheduling by OFFs and increase the cost of CSLs to ar- also finds that integrated logistics is still affecting freight tariffs.
range rescheduling inland transport services. Therefore, accuracy However, timely delivery was not considered in the analysis of
of delivery time is considered the primary consequence factor. previous literature.
There are both unidirectional and bidirectional relationships
between ‘timely delivery,’ ‘integrated logistics,’ ‘transit time’ POSTSCRIPT
and ‘customs clearance efficiency’. These relationships indicate This paper is a revised and expanded version of a paper en-
that ‘timely delivery’ is affected by the aforementioned factors. titled ‘key influential factors of maritime service for ocean freight
Thus OFFs will consider ‘integrated logistics,’ ‘transit time,’ forwarders selecting container shipping lines’ presented at inter-
‘customs clearance efficiency,’ and ‘transport reliability’ when national conference on global integration of economies and con-
considering ‘timely delivery’. Those factors indicate that the nectivity development, 31 August to 1 September 2015, Soochow
services scope and quality standards provided by CSLs, together University, Taipei, Taiwan.
with all the other factors will affect the acceptance of ‘timely
delivery’ for OFFs. As for ‘freight tariffs’ and ‘transport reli-
ACKNOWLEDGMENTS 
ability,’ they will be affected by ‘integrated logistics,’ indicating
that OFFs who are considering ‘freight tariffs’ will also consider The authors thank the Ministry of Science and Technology,
‘integrated logistics’ and ‘transport reliability’. Therefore, CSLs Taiwan, R.O.C., for providing financial supports (MOST 104-
should not only be more flexible in their ‘freight tariffs,’ but 2410-H-019-011 and 105-2410-H-019-010).
 T.-C. Ho et al.: Key Factors for OFFs Selecting CSLs Using RDEMATEL 309

REFERENCES  destination competitiveness: a comparison among six Asian cities. Journal

of Hospitality and Tourism 8(3), 235-252.
Brooks, M. R. (1995). Understanding the ocean container market: a seven Linstone, H. A. and M. Turoff (1975). The Delphi Method: Techniques and Ap-
country study. Maritime Policy and Management 22(1), 39-49. plications. New York, MA: Addison-Wesley.
Brooks, M. R. (1991). Assessment of the ocean carrier decision environment: Lirn, T. C., H. A. Tanooulu, M. J. Beynon and A. K. C. Beresford (2004). An
a longitudinal study. Journal of the Transportation Research Forum 31(2), application of ahp on transshipment port selection: a global perspective.
219-229. Maritime Economics and Logistics 6(1), 70-91.
Brooks, M. R. (1990). Ocean carrier selection criteria in a new environment. Liu, C. Z. and C. L. Lin (2005). A study on leisure agricultural development of
The Logistics and Transportation Reviews 26(4), 339-355. Tai-an Village in Houli Township - An Application of the DEMATEL Method.
Brooks, M. R. (1985). An alternative theoretical approach to the evaluation of Journal of Agricultural and Forest Meteorology 54(4), 263-282.
liner shipping, Part II: Choice Criteria. Maritime Policy and Management McGinnis, M. A. (1990). The relative importance of cost and service in freight
12(2), 145-155. transportation choice: before and after Deregulation. Transportation Journal
Brooks, M. R. and V. Trifts (2008). Short sea shipping in North America: Under- 30(1), 12-19.
standing the requirements of atlantic Canadian shippers. Maritime Policy Mentzer, J. T., D. J. Flint and J. L. Kent (1999). Developing a logistics service
and Management 35(2), 145-158. quality scale. Journal of Business Logistics 20(1), 9-32.
Chen, K. K., C. T. Chang and C. S. Lai (2009). Service quality gaps of business Moeller, G. and E. Shafer (1987). The Delphi Technique: A Tool for Long-
customers in the shipping industry. Transportation Research Part E 45(1), range Tourism and Travel Planning, In Travel, Tourism, and Hospitality
222-237. Research-A Handbook for Managers and Researchers, J. R. Brent Ritchie
Chen, W. L. and T. H. Chen (2011). Application of MDHP method to establish and Charles R., Golden, Eds. New York.
apparel design indices. Journal of Humanities and Social Sciences 7(1), Murphy, P. R. and P. K. Hall (1995). The relative importance of cost and service
49-59. in freight transportation choice before and after Deregulation: an update.
Chung, C. C., Y. S. Chung and A. N. Tai (2011). An evaluation of key service Transportation Journal 35(1), 30-39.
attributes of ocean container carriers from the ocean freight forwarderʼs per- Murry, J. W. and J. O. Hammons (1995). Delphi: a versatile methodology for
spective. Journal of the Eastern Asia Society for Transportation Studies 9, conducting qualitative research. The Review of Higher Education 18(4),
605-620. 423-436.
Collison, F. M. (1984). Market segments for marine liner service. Transportation Notteboom, T. E. (2006). The Time Factor in Liner Shipping Services. Maritime
Journal 24(2), 40-54. Economics and Logistics 8(1), 19-39.
Douglas, V. M., T. J. Page, S. B. Keller and J. Ozments (2006). Determining im- Penaloza, E., M. Brooks and S. Marche (2007). Case study analysis of the im-
portant carrier attributes: a fresh perspective using the theory of reasoned pacts of electronic commerce on the strategic management of container
action. Transportation Journal 45(4), 7-19. shipping companies. Maritime Policy and Management 34(1), 37-45.
Faherty, V. (1979). Continuing social work education: results of a delphi survey. Rogerson, S., D. Andersson and M. I. Johansson(2014). Influence of context on
Journal of Education for Social Work 15(1), 12-19. the purchasing process for freight transport services. International Journal
Heaver, T. D. (2001). The evolving roles of shipping lines in international of Logistics 17(3), 232-248.
logistics. International Journal of Maritime Economics 4(3), 210-230. Rowe, G. and G. Wright (1999). The delphi technique as a forecasting tool:
Hsu, P. F. (2010). Applying the ANP model for selecting the optimal location issues and analysis. International Journal of Forecasting 15(4), 353-375.
for an international business office center in China. Asia Pacific Manage- Salleh, A. L. (2007). worldwide sourcing practice of malaysian electrical and
ment Review 15(1), 27-41. electronics companies. The Business Review Cambridge 8(2), 61-67.
Hsu, C. W., T. C. Kuo, S. H. Chen and A. H. Hu (2013). Using DEMATEL to de- Seyed-Hosseini, S. M., N. Safaei and M. J. Asgharpour (2006). Reprioritization
velop a carbon management model of supplier selection in green supply of failures in a system failure mode and effects analysis by decision making
Chain management. Journal of Cleaner Production 56(1), 164-172. trial and evaluation laboratory technique. Reliability Engineering and System
Huang, S. T. (2014). Customer satisfaction assessment of fourth party logistics Safety 91(8), 872-881.
service providers by using quality function deployment. The International Shry, F. Y. and J. H. Chu (2005). Modeling bargaining strategies between
Journal of Transport and Logistics 41(30), 1-9. forwarders and maritime carriers. Transportation Planning Journal 34(1),
Kannan, V. (2010). Benchmarking the service quality of ocean container car- 119-144.
riers using AHP. Benchmarking: An International Journal 17(5), 637-656. Slack, B. (1985). Containerization, inter-port competition and port selection.
Kannan, V., S. K. Bose and N.G. Kannan (2011). An evaluation of ocean container Maritime Policy and Management 12(4), 293-303.
carrier selection criteria: an Indian shipper’s perspective. Management Re- Tiwari, P., H. Itoh and M. Doi (2003). Shippers’ port and carrier selection be-
search Review 34(7), 754-772. havior in China: a discrete choice analysis. Maritime Economics and Logistics
Krapfel, R. E. and J. T. Mentzer (1982). Shippers transportation choice processes 5(1), 23-39.
under deregulation. Industrial Marketing Management 11(3), 117-124. Tongzon, J. L. (2009). Port choice and freight forwarders. Transportation Re-
Lauterborn, R. (1990). New marketing litany: 4p’s passé; c-words take over. search Part E 45(1), 186-195.
Advertising Age 10(1), 26. Tzeng, G. H., C. H. Chiang and C. W. Li (2007). Evaluating intertwined effects
Lee, H. S., G. H. Tzeng, W. C. Yeih and Y. J. Wang (2013). Revised DEMATEL: in e-learning programs: a novel hybrid MCDM model based on factor ana-
resolving the infeasibility of DEMATEL. Applied Mathematical Modelling lysis and DEMATEL. Expert Systems with Application 32(4), 1028-1044.
37(10-11), 6746-6757. Vernimmen, B., W. Dullaert and S. Engelen (2007). Schedule unreliability in liner
Lee, Y. S., J. C. Huang and Y. S. Hsu (2008). Using modified delphi method to shipping: origins and consequences for the hinterland supply Chain. Mari-
explore the competition strategy for software companies of Taiwan. Journal time Economics and Logistics 9(3), 193-213.
of Informatics and Electronics 3(1), 39-50. Wen, C. H. and W. W. Lin (2016). Customer segmentation of freight forwarders
Liang, G. S., J. F. Ding and I. W. Liu (2007). Using conjoint analysis on the eva- and impacts on the competitive positioning of ocean carriers in the Taiwan-
luation of brand equity for liner service companies: an empirical study of Southern China trade lane. Maritime Policy and Management 43(4), 420-435.
four international carriers. Journal of the Chinese Institute of Transportation Wen, C. H. and J. Y. Huang (2007). A discrete choice model of ocean carrier choice.
19(4), 333-362. Journal of the Eastern Asia Society for Transportation Studies 7, 795-807.
Liao, Y. H., H. A. Lu, S. M. Tsao and K. H. Liao (2004). Ship scheduling for Wong, P. C., H. Yan and C. Bamford (2008). Evaluation of factors for carrier
container liner. Journal of the Chinese Institute of Transportation 16(3), selection in the China Pearl River Delta. Maritime Policy and Management
203-226. 35(1), 27-52.
Lin, F. Y., C. L. Huh and J. J. Yen (2011). Exploring the international exhibition Yang, C. C., H. H. Tai and W. H. Chiu (2014). Factors influencing container
310 Journal of Marine Science and Technology, Vol. 25, No. 3 (2017 )

carriers’ use of coastal shipping. Maritime Policy and Management 41(2), Yeung, C. L. (2006). The impact of third-party logistics performance on the
192-208. logistics and export performance of users: an empirical study. Maritime
Yang, S. C. (2013). Choosing containership registration with revised DEMATEL. Economics and Logistics 8(2), 121-139.
Ph.D. Dissertation, Institute of Shipping and Transportation Management, Zacharia, Z. and J. Mentzer (2004). Logistics salience in a changing environment.
National Taiwan Ocean University, Taiwan, Republic of China. Journal of Business Logistics 25(1), 187-210.
Yen, J. R. and S. M. Chen (2004). Modeling the shippers’ behavior in short-sea Zsidisin, G. A., M. D. Voss and M. Schlosser (2007). Shipper-carrier relationships
routes. Maritime Quarterly 13(2), 73-96. and their effect on carrier performance. Transportation Journal 46(2), 5-18.