STRATEGIES FOR IMPROVING ACCREDITATION
PERFORMANCE IN HIGHER EDUCATION INSTITUTION
Kua-Hsin Peng
1
and Gwo-Hshiung Tzeng
1,2
1
Institute of Project Management, Kainan University, No. 1, Kainan Road, Luchu, Taoyuan 338, Taiwan
2
Institute of Management of Technology, National Chiao Tung University, 1001 Ta-Hsueh Road, Hsinchu 300, Taiwan
Keywords: Accreditation, Performance, Improvement Strategy, MCDM (Multiple Criteria Decision Making),
DEMATEL, DANP (DEMATEL-based ANP), VIKOR.
Abstract: Numerous studies has focused on exploring input and output indicators of accreditation system; assessment
quality assurance and accreditation of higher education; reviewing the status of quality assurance and
accreditation system within higher education. However, few studies have explored strategies for improving
institutional accreditation performance in higher education, and preventing decision makers from obtaining
valuable cues for making accurate decisions to improve institutional accreditation performance to increase
the logical thinking, reasoning ability and work competitiveness of graduate students. Therefore, the
purpose of this study was to explore strategies for improving institutional accreditation performance using a
new hybrid MCDM model combined with DANP (DEMATEL-based ANP). An empirical case was to
demonstrate the effectiveness of the proposed model for evaluating institutional accreditation performance
to identify institutional performance gaps and explore strategies for improving accreditation based on the
influential relation map. Decision makers should increase the priority of the cause criteria in advance, to
successfully improve institutional accreditation performance to achieve the aspiration levels and increase
competiveness of students.
1 INTRODUCTION
Currently, the number and size of higher education
institutions and the diversity of programs offered is
significantly increasing (Aqlan et al., 2010).
However, in the contemporary changing and
uncertain world, all higher education institutions
should respond favorably to social needs
(Yarmohammadian et al., 2011). Evaluation thus is
one of the strongest tools for strategic development
in higher education environments (Saad, 2001).
Professional higher education planners can use
evaluation to identify their strengths and weaknesses
and assume responsibility for educational needs at
the national and global levels, and to continuously
improve educational process and program quality
(Yarmohammadian et al., 2011); (Wild, 1995);
(ForoughiAbari et al., 2004). Consequently, interest
is growing in establishing quality assurance and
accreditation systems in higher education (Anaam et
al., 2009).
Accreditation involves external quality review
created and used by higher education authorities to
assure and improve quality in colleges, universities
and programs (Eaton, 2006). The accreditation
process provides colleges and universities with an
opportunity for reflection, honest assessment of
strengths and weaknesses, and the development of
strategies for continued improvement. Additionally,
the accreditation process aims to guarantee the
quality of educational programs by ensuring that
graduates have acquired the necessary knowledge,
skills and attitude to work successfully in their
chosen profession, and that the educational
programs/institutions are satisfactory to various
stakeholders. Consequently, the main influences of
an accreditation system include encouraging quality
improvement initiatives by institutions, improving
student enrolment quantity and quality, helping
institutions attract and retain better quality faculty,
helping institutions secure funding, enhancing
graduate employability, facilitating trans-national
recognition of degrees and mobility of graduates and
professionals, motivating faculty to participate
actively in academic and related institutional
/departmental activities, helping create a sound and
challenging academic environment in institutions
211
Peng K. and Tzeng G..
STRATEGIES FOR IMPROVING ACCREDITATION PERFORMANCE IN HIGHER EDUCATION INSTITUTION.
DOI: 10.5220/0003900902110221
In Proceedings of the 4th International Conference on Computer Supported Education (CSEDU-2012), pages 211-221
ISBN: 978-989-8565-07-5
Copyright
c
 2012 SCITEPRESS (Science and Technology Publications, Lda.)

and contributing to national social and economic
development by producing high quality technical
manpower (Campbell et al., 2002); (Anthony, 2004);
(UNESCO, 2007); (Prasad and Bhar, 2010).
Extensive literature has focused on exploring input
and output indicators of accreditation system
(Cavaller, 2011); assessment quality assurance and
accreditation of higher education (Yarmohammadian
et al., 2011); (Aqlan et al., 2010); and reviewing the
status of quality assurance and accreditation system
within higher education (Anaam et al., 2009).
However, few studies have explored strategies to
improve institutional accreditation performance for
increasing higher education quality, and preventing
decision makers from obtaining valuable cues for
making accurate decisions to improve institutional
accreditation performance to increase the logical
thinking, reasoning ability and work competitiveness
of graduate students.
MCDM (Multiple Criteria Decision Making) is
scientifically analytical method that can help
decision-makers selects the best alternative among
multiple criteria (Tsaur et al., 1997); (Wang and Lee,
2009). Consequently, a hybrid MCDM method has
been developed and is widely used in numerous
fields. Tsaur et al. (1997) applied a Fuzzy MCDM to
evaluate tourist risk. Furthermore, Ou Yang et al.,
(2008) combined DEMATEL technique and ANP to
solve the dependence and feedback problems to suit
the real world. Furthermore, Hung et al., (2011) used
the hybrid MCDM model to solve this knowledge
management systems adoption problem.
Additionally, Kuan et al., (2011) used the hybrid
MCDM model to assess the total performance of the
new product development (NPD) process.
Additionally, Yang and Tzeng (2011) demonstrated
how the DEMATEL technique clarified the
direct/indirect influential relationship of criteria.
Decision-makers can use the influential relation map
to identify the key criterion for improving
institutional accreditation performance.
Based on the above discussion, this study
attempts to explore the strategy for improving
institutional accreditation performance using a new
hybrid MCDM model that is combined with DANP
(DEMATEL-based ANP). An empirical case is also
presented to demonstrate the effectiveness of a new
hybrid MCDM model combining DANP, and
VIKOR is used to assess institutional accreditation
performance to identify the performance gaps and
explore strategies for improving institutional
accreditation performance based on the influential
relation map by DEAATEL technique. The best
improvement strategy for promoting institutional
accreditation performance to reduce the gaps in each
criterion and achieve the aspiration levels can then
be obtained and implemented.
The remainder of this paper is organized as
follows. Section 2 develops a new hybrid MCDM
model for exploring institutional accreditation
performance improvement strategy. Section 3 then
presents an empirical case analysis of institutional
accreditation performance to illustrate the proposed
model. Finally, the last section presents conclusions.
2 METHODOLOGY
This Section comprises four parts: the first part
describes the data collection; the second part
presents the DEMATEL technique for building a
network relationship; the third part calculates the
influential weights using DANP (DEMATEL-based
ANP); finally, the last part uses VIKOR to evaluate
total accreditation performance.
2.1 Data Collection
Table 1 lists criteria for evaluating accreditation
performance based on Higher Education Evaluation
& Accreditation Council of Taiwan (HEEACT). The
survey targeted professors of university. First, this
study used a four-point scale ranging from 0 (no
influence) to 4 (very high influence) to identify the
evaluation criteria and their influence on one
another. Ten professors were then asked to assess
the influence of the criteria on one another, and the
consensus rates of the dimensions and criteria were
97.44% and 97.05% (both exceeding 97%),
respectively. Finally, ten experts are asked to
evaluate the level of importance and performance for
each criterion. Furthermore, this study used VIKOR
method to assess total accreditation performance,
identify the gaps in performance, and explore the
strategy for improving accreditation performance
based on the influential relation map.
2.2 DEMATEL Technique for
Establishing a Network
Relationship
DEMATEL is an analytical technique for building a
structural model (see Appendix A, A1).
DEMATEL is mainly used to solve complex
problems to clarify their essential nature.
DEMATEL uses matrix and related mathematical
theories (Boolean operation) to calculate the cause
CSEDU2012-4thInternationalConferenceonComputerSupportedEducation
212

and effect relationships involved in each element.
This technique is widely used to solve various
complex studies, and particularly to understand
complex problem structures and provide viable
problem-solving methods (Tzeng et al., 2007).
Table 1: Evaluation criteria.
Dimensions Criteria
Settings
goals and
features
and self-
improvement
(
1
D
)
Cognition of teachers and students regarding
educational goals (
1
C
)
Teaching and learning activities reflect the goals of
the educational institutions (
2
C
)
Operations of the self-accreditation system(
3
C
)
Effectiveness of the self-improvement system (
4
C
)
Course
design and
teaching
(
2
D
)
Operations of the course planning system (
5
C
)
Teacher quality and quantity meet student learning
and teaching needs (
6
C
)
Teachers teach students according to the syllabus
(
7
C
)
Institutions emphasize teacher professional
development and teaching improvement (
8
C
)
Student
learning and
guidance
(
3
D
)
Teaching meets student learning needs (
9
C
)
Institution teaching resources meet students learning
needs (
10
C
)
Institution provides student counselling, life
coaching, career counselling, etc. (
11
C
)
Teachers provide students with a fixed interview
time (
12
C
)
Institutions respond to student comments (
13
C
)
Students interact with advisors (
14
C
)
Research
and
professional
performance
(
4
D
)
Research results and professional performance of
teachers (
15
C
)
Teachers obtained research project grants (
16
C
)
Effectiveness of teacher participation in social
services
17
()C
Performance of student learning outcomes (
18
C
)
Graduate
performance
(
5
D
)
Institution established an effective channel for
tracking graduate performance (
19
C
)
Job directors perceive satisfaction with graduates
(
20
C
)
Graduate jobs, salary and other achievements (
21
C
)
Source: Higher Education Evaluation & Accreditation Council of
Taiwan (HEEACT).
The DEMATEL technique comprises five steps.
The first step is to confirm the system has n
elements and develop the evaluation scale, using a
pair-wise of dimensions to perform the comparison,
and also using the measuring scale 0, 1, 2, 3, 4,
representing complete no influence (0), low
influence (1), medium influence (2), high influence
(3), and extremely high influence (4) as natural
language by pair-wise comparison. The second step
is to calculate the initial matrix to directly obtain
influential matrix (Lin and Tzeng, 2009); (Chen et
al., 2010). The third step is to normalize the matrix
such that at least one column or row, but not all,
sums to one. The forth step is to obtain the total
influence matrix. Moreover, the fifth step is to
obtain prominence and relation to build the
influential relation map. DEMATEL is based on the
concept of influential relation map, which can
distinguish the direct/indirect influential relationship
of the criteria, allowing decision-makers to identify
the key criterion for developing strategies for
improving accreditation performance in higher
education of this study.
2.3 Finding the Influential Weights
using the DANP
This study not only uses the DEMATEL technique
to confirm the interactive relationship among the
various dimensions/criteria, but also seeks the most
accurate influential weights. This study found that
ANP can serve this purpose. This study used the
basic concept of ANP (Saaty, 1996), which
eliminates the limitations of Analytic Hierarchy
Process (AHP) and is applied to solve nonlinear and
complex network relations (Saaty, 1996). ANP is
intended to solve interdependence and feedback
problems of criteria. This study thus applies the
characteristics of influential weights ANP and
combines them with DEMATEL (call DANP,
DEMATEL-based ANP) to solve these kind of
problems based on the basic concept of ANP (see
Appendix A, A2). This approach yields more
practical results.
2.4 Evaluating Competitiveness Gaps
using VIKOR
Opricovic and Tzeng (2004) proposed the
compromise ranking method (VIKOR) as a suitable
technique for implementation within MCDM
(Opricovic, 1998); (Tzeng et al., 2005); (Opricovic
and Tzeng, 2004; 2007). VIKOR uses the class
distance function (Yu, 1973) based on the concept of
the Positive-ideal (or we adopt the Aspiration level)
solution and Negative-ideal (or we adopt the Worst
level) solution and puts the results in order. For
normalized class distance function it is better to be
near the positive-ideal point (the aspiration level)
and far from the negative-ideal point (the worst
value) for normalized class distance function (Lee et
al., 2009); (Ho et al., 2011). VIKOR comprises the
following steps: The first step is to check the best
and worst values of the assessment criteria. The
STRATEGIESFORIMPROVINGACCREDITATIONPERFORMANCEINHIGHEREDUCATIONINSTITUTION
213

second step is to calculate the mean of group utility
based on the sum of all individual-criterion regret
(i.e., average overall performance gaps, as well as
those for each dimension, and for each criterion; as
well as strategies for reducing these gaps), and
calculate the maximal regret of an individual-
criterion for improvement priority, both overall and
for each dimension. The third step is to obtain the
comprehensive/integrating indicators and sorting
results provided to the decision-maker to implement
improvement strategies and reduce competitiveness
gaps in overall and each dimensional performance
(see Appendix B).
3 AN EMPIRICAL CASE OF
TAIWAN
This section presents an empirical case involving
Taiwan to explore strategies for improving
accreditation performance based on a new hybrid
MCDM model. The contents include background
and problem description, analysis results of
accreditation performance, and measurement of the
cause and effect relationships among the evaluation
criteria; this framework is then used to identify
institutional accreditation performance gaps and
explore strategies for improving accreditation based
on the influential relation map.
3.1 Background and Problem
Description
The number of higher education institutions in
Taiwan has recently increased rapidly, thus, the
greatest challenge facing higher education in Taiwan
is how to assure quality and competitiveness in the
current era of globalization (Hou and Morse, 2009).
Consequently, under the “University Law” revised
in 2005, all Taiwanese universities and colleges are
obliged to undergo regular assessments relating to
standards and procedures by accrediting agencies
chartered by the Ministry of Education (Hou and
Morse, 2009). In the same year, the HEEACT was
officially established and began to conduct
evaluations of Taiwanese higher education programs
in 2006 (HEEACT, 2008). However, Sadlak (2010)
presented that universities and other higher
education institutions are rightly seen as
powerhouses and nurseries that are essential for
economic development and global competitiveness.
Given this situation, the best method of exploring
strategies for improving accreditation to increase the
quality of higher education in Taiwan has become
important, and thus decision-makers can obtain
valuable cues for making decisions to improve
performance to the desired level.
3.2 Analysis of Results
The DEMATEL technique is used to construct an
NRM (network relation map) that illustrates
influential networks of five dimensions with 21
criteria of accreditation. Based on DEMATEL
technique, this study obtained the total influence
matrix T of the dimensions and criteria, as shown in
Tables 2 to 4. According to the influential
relation
()r-d
ij
, “Settings goals and features and self-
improvement (
1
D
)” is the highest degree of an
impact relationship that affects other dimensions
directly. Otherwise, “Graduate performance (
5
D
)” is
the most vulnerable to impact.
Table 3 lists all the criteria of the influential
relation with each criterion. Table 4 lists the
relationship between the extents of the direct or
indirect impacts and compares them with other
criteria. “Effectiveness of self-improvement system
(
4
C
)” is the most important consideration criteria;
additionally, “Effectiveness of teacher participation
in social services (
17
C
)” is the influence of all
criteria in the least degree of other criteria.
Furthermore, Table 4 shows that “Institution respond
to student comments (
13
C
)” is the highest degree of
influential relationship in all the criteria. Otherwise,
“Graduate jobs, salary, achievements (
21
C
)” is the
most vulnerable to impact of criteria that compare
with other criteria.
Table 2: Total influence matrix of
D
T
and the sum of the
influences on the dimensions.
1
D
2
D
3
D
4
D
5
D
i
r
i
d
ii
rd+
ii
rd
−
1
D
1.942 2.277 2.248 2.119 2.291 10.877 9.683 20.560 1.193
2
D
2.085 2.002 2.197 2.050 2.244 10.577 10.266 20.843 0.311
3
D
1.950 2.078 1.880 1.936 2.127 9.970 10.227 20.196 -0.257
4
D
1.848 1.943 1.920 1.655 2.005 9.370 9.587 18.957 -0.217
5
D
1.860 1.967 1.983 1.826 1.832 9.469 10.499 19.968 -1.031
Note: average gap =
1
11
1
100%
(1)
nn
nn
ij ij
n
ij
ij
gg
nn
g
−
==
−
×
−
∑∑
=2.56% <
5%,
n denotes the samples of 10 experts and the consensus rate is
97.44 %.
This study not only uses DEMATEL technique
to confirm the interfering relationship with the
criteria, but also expects to obtain the most accurate
influential weights. ANP is applied to solve the
interdependence and feedback problems of criteria.
CSEDU2012-4thInternationalConferenceonComputerSupportedEducation
214

Table 3: The total influence matrix of
C
T
for criteria.
1
C
2
C
3
C
4
C
5
C
6
C
7
C
8
C
9
C
10
C
11
C
12
C
13
C
14
C
15
C
16
C
17
C
18
C
19
C
20
C
21
C
1
C
0.258 0.330 0.321 0.339 0.320 0.320 0.276 0.337 0.314 0.334 0.285 0.263 0.269 0.303 0.275 0.237 0.224 0.288 0.258 0.264 0.264
2
C
0.310 0.281 0.328 0.350 0.324 0.325 0.280 0.338 0.311 0.329 0.284 0.265 0.271 0.301 0.279 0.245 0.228 0.290 0.258 0.272 0.272
3
C
0.313 0.339 0.278 0.348 0.322 0.321 0.285 0.342 0.316 0.343 0.291 0.277 0.280 0.307 0.286 0.250 0.234 0.290 0.269 0.269 0.267
4
C
0.321 0.358 0.354 0.314 0.342 0.351 0.304 0.363 0.332 0.354 0.306 0.296 0.299 0.363 0.299 0.269 0.244 0.309 0.291 0.293 0.293
5
C
0.296 0.324 0.310 0.333 0.259 0.307 0.274 0.323 0.299 0.316 0.265 0.253 0.262 0.287 0.268 0.236 0.211 0.283 0.240 0.257 0.260
6
C
0.315 0.345 0.329 0.353 0.324 0.283 0.297 0.349 0.324 0.339 0.292 0.276 0.285 0.314 0.281 0.255 0.231 0.293 0.264 0.280 0.280
7
C
0.253 0.277 0.271 0.286 0.268 0.272 0.196 0.279 0.258 0.273 0.238 0.230 0.227 0.252 0.226 0.209 0.188 0.240 0.208 0.218 0.220
8
C
0.307 0.327 0.322 0.340 0.315 0.322 0.277 0.283 0.305 0.329 0.283 0.268 0.280 0.310 0.286 0.259 0.231 0.286 0.253 0.267 0.267
9
C
0.325 0.347 0.342 0.364 0.344 0.347 0.292 0.350 0.278 0.353 0.306 0.289 0.295 0.320 0.290 0.259 0.233 0.305 0.278 0.289 0.292
10
C
0.312 0.337 0.328 0.348 0.332 0.337 0.287 0.340 0.319 0.290 0.305 0.288 0.294 0.319 0.282 0.247 0.231 0.302 0.269 0.279 0.281
11
C
0.285 0.311 0.302 0.319 0.303 0.309 0.259 0.305 0.292 0.323 0.230 0.262 0.264 0.296 0.257 0.229 0.208 0.278 0.248 0.255 0.255
12
C
0.246 0.263 0.249 0.264 0.243 0.256 0.222 0.261 0.249 0.269 0.238 0.183 0.233 0.255 0.207 0.185 0.171 0.229 0.219 0.218 0.221
13
C
0.304 0.324 0.311 0.328 0.308 0.318 0.275 0.318 0.305 0.326 0.286 0.271 0.228 0.299 0.251 0.223 0.208 0.276 0.265 0.258 0.259
14
C
0.273 0.300 0.287 0.305 0.286 0.288 0.249 0.302 0.281 0.297 0.258 0.246 0.252 0.240 0.263 0.241 0.219 0.279 0.245 0.248 0.246
15
C
0.281 0.294 0.294 0.306 0.284 0.299 0.245 0.313 0.277 0.296 0.258 0.234 0.240 0.284 0.218 0.252 0.230 0.281 0.233 0.250 0.244
16
C
0.242 0.258 0.259 0.272 0.250 0.266 0.218 0.279 0.244 0.262 0.227 0.210 0.213 0.255 0.245 0.172 0.200 0.247 0.211 0.222 0.218
17
C
0.215 0.227 0.225 0.237 0.216 0.229 0.191 0.238 0.212 0.225 0.199 0.182 0.189 0.217 0.208 0.188 0.137 0.215 0.192 0.196 0.194
18
C
0.306 0.320 0.313 0.329 0.312 0.316 0.268 0.325 0.302 0.318 0.282 0.263 0.269 0.305 0.282 0.244 0.227 0.245 0.263 0.283 0.283
19
C
0.244 0.271 0.264 0.278 0.261 0.256 0.219 0.263 0.248 0.263 0.233 0.215 0.230 0.249 0.226 0.201 0.181 0.237 0.182 0.228 0.224
20
C
0.246 0.271 0.263 0.280 0.262 0.256 0.218 0.263 0.244 0.262 0.233 0.212 0.222 0.240 0.227 0.198 0.184 0.231 0.224 0.188 0.244
21
C
0.216 0.233 0.230 0.241 0.223 0.228 0.192 0.230 0.215 0.234 0.206 0.185 0.202 0.224 0.202 0.177 0.162 0.214 0.207 0.212 0.167
Note: average gap =
1
11
1
100%
(1)
nn
nn
ij ij
n
ij
ij
gg
nn
g
−
==
−
×
−
∑∑
= 2.95% < 5%, n denotes the samples of 10 experts and the consensus rate is 97.05%.
Therefore, this study builds the accreditation
performance assessment model using DEMATEL
technique, which is combined with the DANP
(DEMATEL-based ANP) model to obtain the
influential weights of each criterion, as shown in
Table 4. Additionally, the critical criteria in
accreditation performance assessment of University
in Taiwan (a business school as example in
accreditation performance assessment) are identified
as “Job directors perceive satisfaction with graduates
20
()C
”, “Graduate jobs, salary and other
achievements
21
()"C
and “Institution established an
effective channel for tracking graduate performance
(
19
C
)”. Furthermore, the influential weights combine
with the DEMATEL technique to assess the priority
of problem-solving based on the performance gaps
identified by VIKOR method and the influential
relation map.
An empirical case involving University in Taiwan
is used to evaluate the total accreditation
performance using the VIKOR method, as listed in
Table 5. The scores of each criterion and the total
average gap
()
k
S
of University in Taiwan are
obtained, using the relative influential weights from
DANP to multiply the gap
()
kj
r . Consequently, this
study obtains the total performance gap of
University in Taiwan based on the scoring value.
Additionally, the comprehensive/integrating
indicator
()
k
R
can be obtained, which value of v can
make decisions by the expert that is defined as
1v
=
,
0.5v
=
, and
0v
=
in this paper. This study obtains
the result of the comprehensive/integrating
indicators
()
k
R
as 0.313 (total average gap), 0.3814
(the majority of criteria), and 0.450 (maximal gap of
priority improvement) representing that a business
school as example in accreditation performance
assessment (HEEACT) must improve the gap of
accreditation performance. Furthermore, the
ministry of education can find the problem-solving
points according to the DEMATEL technique
combined with DANP and VIKOR (called the
hybrid MCDM model).
3.3 Discussions and Implications
This study adopted a new hybrid MCDM model
using the DEMATEL technique combined with
STRATEGIESFORIMPROVINGACCREDITATIONPERFORMANCEINHIGHEREDUCATIONINSTITUTION
215

Table 4: The sum of the effects, weights and rankings of
each criterion.
Criteria
i
r
j
d
ij
rd+
ij
rd−
Degree of importance
(Global weights)
1
D
0.1930 (4)
1
C
6.078 5.864 11.942 0.214 0.0454 (4)
2
C
6.141 6.339 12.480 -0.198 0.0491 (2)
3
C
6.228 6.180 12.408 0.048 0.0479 (3)
4
C
6.656 6.532 13.187 0.124 0.0506 (1)
2
D
0.2043 (2)
5
C
5.863 6.097 11.960 -0.234 0.0519 (3)
6
C
6.309 6.205 12.514 0.104 0.0528 (2)
7
C
5.087 5.322 10.409 -0.235 0.0452 (4)
8
C
6.116 6.400 12.516 -0.285 0.0544 (1)
3
D
0.2035 (3)
9
C
6.495 5.925 12.420 0.570 0.0353 (3)
10
C
6.326 6.335 12.661 -0.010 0.0377 (1)
11
C
5.785 5.505 11.290 0.280 0.0328 (4)
12
C
4.882 5.169 10.051 -0.287 0.0307 (6)
13
C
5.941 5.302 11.243 0.640 0.0316 (5)
14
C
5.604 5.939 11.543 -0.336 0.0354 (2)
4
D
0.1906 (5)
15
C
5.611 5.356 10.967 0.255 0.0508 (2)
16
C
4.968 4.776 9.744 0.192 0.0452 (3)
17
C
4.331 4.383 8.713 -0.052 0.0415 (4)
18
C
6.055 5.616 11.671 0.438 0.0531 (1)
5
D
0.2086 (1)
19
C
4.973 5.078 10.050 -0.105 0.0680 (3)
20
C
4.966 5.243 10.209 -0.277 0.0703 (1)
21
C
4.401 5.249 9.650 -0.848 0.0703 (1)
( ) denotes ranking
DANP (DEMATEL-based ANP) with VIKOR
method to explore the improvement strategies of
accreditation performance in the empirical case of a
business school of University at Taiwan. Figure 1
shows valuable cues for making accurate decisions.
The influential relation map demonstrate that the
degrees of influence among dimensions and criteria.
This study applies the most important and influential
criteria as critical criteria to improve the maximal
gap of accreditation performance. This list of critical
criteria can provide a reference for Taiwanese
ministry of education to develop the improving
strategic to successfully improve institutional
accreditation performance and increase
competiveness of students.
The following recommendations are proposed to
improve institutional accreditation performance of
high education in Taiwan. This system structure
model illustrates that University in Taiwan suffers
Table 5: The performance evaluation of the case study by
VIKOR.
Dimensions
/ Criteria
Local weight
Global weight
(by DANP)
Case study of Taiwan
Score
Gap
()
kj
r
1
D
0.1930(4) 7.19 0.281
1
C
0.2352 0.0454(4) 7.33 0.267
2
C
0.2544 0.0491(2) 7.33 0.267
3
C
0.2482 0.0479(3) 7.25 0.275
4
C
0.2622 0.0506(1) 6.83 0.317
2
D
0.2043(2) 6.63 0.338
5
C
0.2540 0.0519(3) 6.83 0.317
6
C
0.2584 0.0528(2) 7.17 0.283
7
C
0.2212 0.0452(4) 5.92 0.408
8
C
0.2663 0.0544(1) 6.58 0.342
3
D
0.2035(3) 7.15 0.285
9
C
0.1735 0.0353(3) 7.50 0.250
10
C
0.1853 0.0377(1) 7.42 0.258
11
C
0.1612 0.0328(4) 6.92 0.308
12
C
0.1509 0.0307(6) 6.33 0.367
13
C
0.1553 0.0316(5) 7.17 0.283
14
C
0.1740 0.0354(2) 7.58 0.242
4
D
0.1906(5) 6.75 0.325
15
C
0.2665 0.0508(2) 7.50 0.250
16
C
0.2371 0.0452(3) 6.67 0.333
17
C
0.2177 0.0415(4) 5.50 0.450
18
C
0.2786 0.0531(1) 7.33 0.267
5
D
0.2086(1) 6.55 0.344
19
C
0.3260 0.0680(3) 6.58 0.342
20
C
0.3370 0.0703(1) 6.75 0.325
21
C
0.3370 0.0703(1) 6.33 0.367
Total performances -
6.90 -
Total gap
()
k
S
-
- 0.313
CSEDU2012-4thInternationalConferenceonComputerSupportedEducation
216

significant gap in the “Graduate performance
5
()"D
dimensions, making it necessary to pay attention to
the “Settings goals and features and self-
improvement (
1
D
)＂, “Course design and teaching
(
2
()"D
, “Student learning and guidance
3
()"D
,
“Research and professional performance
4
()"D
dimensions for improving accreditation performance
of University in Taiwan.
Furthermore, for improving the settings goals
and features and self-improvement (
1
D
) dimension,
this study finds that the criterion of “Effectiveness of
self-improvement system
4
()"C
prioritizes improving
the maximal performance gap. Figure 1 shows that
the criteria of “Cognition of teachers and students
regarding educational goals
1
()"C
is the most
important and influential criteria, and thus can be
considered the critical criteria for improving
effectiveness of self-improvement system. Thus, the
criteria of “Cognition of teachers and students
regarding educational goals (
1
C
)＂can be considered
the critical criterion for improving the settings goals
and features, self-improvement.
For improving the
course design and teaching
(
2
D
) dimension, this study finds that the criterion of
“Teachers teach student according to syllabus
7
()"C
is the maximal performance gap. Furthermore, the
criteria of “Teacher quality and quantity meet
student learning and teaching needs (
6
C
)＂ is the
most important and influential criteria, and thus can
be considered the critical criteria for improving
teachers according to syllabus to teach student.
Thus, the criteria of “Teacher quality and quantity
meet student learning and teaching needs (
6
C
)＂can
be considered the critical criterion for improving the
course design and teaching.
For improving the
student learning and guidance
(
3
D
) dimension, this study finds that the criterion of
“Teachers provide students with a fixed interview
time (
12
C
)＂ is the maximal performance gap.
Furthermore, the criteria of “Teaching meets student
learning needs
9
()C
＂, “Institution provides student
counselling, life coaching and career counselling etc.
(
11
C
)＂and “Institution respond to student
comments
13
()C
＂ is the most important and
influential criteria, and thus can be considered the
critical criteria for improving teachers provide
students with a fixed interview time. Thus, the
criteria of “Teaching meets student learning needs
9
()C
＂, “Institution provides student counselling, life
coaching and career counselling etc. (
11
C
)＂and
“Institution respond to student comments (
13
C
)＂can
be considered the critical criterion for improving the
student learning and guidance.
For improving the
research and professional
performance
4
()
D
dimension, this study finds that
the criterion of “Effectiveness of teachers
participation in social services
17
()C
＂ is the
maximal performance gap. Furthermore, the criteria
of “Research results and professional performance
of teachers
15
()C
＂, “Teachers obtained research
project grants (
16
C
)＂and “Performance of student
learning outcomes (
18
C
)＂ is the most important and
influential criteria, and thus can be considered the
critical criteria for improving the effectiveness of
teachers to participate in social service. Thus, the
criteria of “Research results and professional
performance of teachers (
15
C
)＂, “Teachers
obtained research project grants (
16
C
)＂and
“Performance of student learning outcomes
(
18
C
)＂can be considered the critical criterion for
improving the research and professional
performance.
4 CONCLUSIONS
This study can help decision-making to improve
accreditation performance. Furthermore, this study
uses the DEMATEL technique to develop cause-
and-effect influential relationships, then, calculates
the weight using DANP. Finally, this study uses
VIKOR method to evaluate total and dimensional
performances, thus contributing to subsequent
research; for example, future studies should evaluate
the effectiveness of implementing the improvement
strategies of accreditation performance.
As noted above, this study can obtain valuable
cues for making accurate decisions. The graduate’s
performance dimensions exhibit a significant
performance gap, and the
“settings goals and
features and self-improvement (
1
D
), course design
and teaching (
2
D
), student learning and guidance
(
3
D
), and research and professional performance
(
4
D
) dimensions may need to be considered to
improve accreditation performance. Furthermore, to
improve the settings goals and features and self-
improvement (
1
D
) dimension, this study finds the
criterion of effectiveness of self-improvement
system is the maximal performance gap. Therefore it
is necessary to improve the priorities of the cause
criteria, namely cognition of teachers and students
STRATEGIESFORIMPROVINGACCREDITATIONPERFORMANCEINHIGHEREDUCATIONINSTITUTION
217

regarding educational goals. To improve the course
design and teaching (
2
D
) dimension, this study finds
that the criterion of teachers according to syllabus to
teach student exhibits the maximal performance gap.
Furthermore, it is necessary to consider the need to
improve quality and quantity of teacher meet student
learning and teaching needs, to enhance course
design and teaching. To improve the student
learning and guidance (
3
D
) dimension, this study
finds that the criterion of teachers provide students
with a fixed interview time exhibits the maximal
performance gap. Furthermore, it is necessary to
consider the need to improve teaching meets student
learning needs, institution provides student
counselling, life coaching and career counselling etc.
and institution respond to student comments. To
improve the research and professional performance
(
4
D
) dimension, this study finds that the criterion of
the effectiveness of teachers participation in social
services exhibits the maximal performance gap.
Furthermore, it is necessary to consider the need to
improve research results and professional
performance of teachers, teachers obtained research
project grants and performance of student learning
outcomes.
Based on the above, the ministry of education
should increase its prioritization of the cause criteria,
allowing it to successfully improve accreditation
performance to achieve the aspired/desired levels
and increase competiveness of students.
Figure 1: The influential relation map of each dimension
and criteria.
REFERENCES
Anaam, M., Alhammadi, A. O., Kwairan, A. A., 2009.
The status of quality assurance and accreditation
systems within higher education institutions in the
republic of Yemen, Quality in Higher Education,
15(1), 51-61.
Anthony, S., 2004. External quality assurance in Indian
higher education: Development of a decade. Quality in
Higher Education, 10 (2), 115–127.
Aqlan, F., Al-Araidah, O., Al-Hawari, T., 2010. Quality
assurance and accreditation of engineering education
in Jordan, European Journal of Engineering
Education, 35 (3), 311–323.
Campbell, C., Rozsnyai, C., 2002. Quality assurance and
the development of course programmes, Papers on
Higher Education Regional University Network on
Governance and Management in Higher Education in
South East Europe, UNESCO, Bucharest.
Cavaller, V., 2011. Protfolios for entrepreneurship and
self-evaluation of higher education institutions.
Procedia Social and Behavioral Sciences, 12, 19-23.
Chen, Y. C., Lien, H. P., Tzeng, G. H., 2010. Measures
and evaluation for environment watershed plans using
a novel hybrid MCDM model, Expert Systems with
Applications, 37(2), 926-938.
Eaton, J. S., 2006. An Overview of U.S. Accreditation,
Council for Higher Education Accreditation.
Foroughi Abari, A. A.,Yarmohammadian, M. H., Toroqi,
J., 2004. Effectiveness in Higher Education,
Encyclopedia of Higher Education, Ministry of
Science, Research, and Technology, Tehran, Iran.
HEEACT, 2008. 2007 HEEACT annual report. Taipei:
Higher Education Evaluation & Accreditation Council
of Taiwan.
Ho, W. R. J., Tsai, C. L., Tzeng, G. H., Fang, S. K., 2011.
Combined DEMATEL technique with a novel MCDM
model for exploring portfolio selection based on
CAPM, Expert Systems with Applications, 38(1), 16-
25.
Hou, Y. C., Morse, R., 2009. Quality assurance and
excellence in Taiwan higher education- an analysis of
three major Taiwan college rankings, Evaluation in
Higher Education, 3(2), 45-72.
Hung, Y. H., Chou, S. C. T., Tzeng, G. H., 2011.
Knowledge management adoption and assessment for
SMEs by a novel MCDM approach, Decision Support
Systems (forthcoming). Available online 5 February
2011.
Kuan, M. J., Hsiang, C. C., Tzeng, G. H., 2011. Probing
the innovative quality system for NPD process based
on combining DANP with MCDM model,
International Journal of Innovative Computing.
Information and Control (special issue) (forthcoming).
Lee, W. S., Tzeng, G. H., Cheng, C. M., 2009. Using
novel MCDM methods based on fama-French three-
factor model for probing the stock selection. APIEMS,
Dec. 14-16: 1460-1474.
Lin, C. L., Tzeng, G. H., 2009. A value-created system of
science (technology) park by using DEMATEL.
Expert Systems with Applications, 36(6), 9683-9697.
Opricovic, S., Tzeng, G. H., 2004. Compromise solution
by MCDM methods: A comparative analysis of
CSEDU2012-4thInternationalConferenceonComputerSupportedEducation
218

VIKOR and TOPSIS, European Journal of
Operational Research, 156(2), 445-455.
Opricovic, S., 1998. Multicriteria Optimization of Civil
Engineering Systems: Faculty of Civil Engineering,
Belgrade.
Opricovic, S., Tzeng, G. H., 2003. Fuzzy multicriteria
model for post-earthquake land-use planning, Natural
Hazards Review, 4(2), 59-64.
Opricovic, S., Tzeng, G. H., 2007. Extended VIKOR
method in comparison with outranking methods.
European Journal of Operational Research, 178(2),
514-529.
Ou Yang, Y. P., Shieh, H. M., Leu, J. D., Tzeng, G. H.,
2008. A novel hybrid MCDM model combined with
DEMATEL and ANP with applications, International
Journal of Operations Research, 5(3), 1-9.
Prasad, G., Bhar, C., 2010. Accreditation system for
technical education programmes in India: A critical
review, European Journal of Engineering Education,
35 (2), 187–213
Saad, G., 2001. Strategic performance evaluation:
descriptive and prescriptive analysis, Industrial
Management & Data Systems, 101, 390-399.
Saaty, T. L., 1996. Decision Making with Dependence and
Feedback: The Analytic Network Process. RWS
Publications, Pittsburgh, PA.
Sadlak, J., 2010. Quality challenge in a changing
landscape of higher education: Place and impact of
academic rankings, Evaluation in Higher Education,
4(1), 1-12.
Tsaur, S. H., Tzeng, G. H., Wang, K. C., 1997. Evaluating
tourist risks from fuzzy perspectives, Annals of
Tourism Research, 24(4), 796-812.
Tzeng, G. H., Chiang, C. H., Li, C. W., 2007. Evaluating
intertwined effects in e-learning programs: A novel
hybrid MCDM model based on factor analysis and
DEMATEL, Expert Systems with Applications, 32(4),
1028-1044.
Tzeng, G. H., Lin, C. W., Opricovic, S., 2005. Multi-
criteria analysis of alternative-fuel buses for public
transportation. Energy Policy, 33(11), 1373-1383.
UNESCO, External quality assurance: Options for higher
education managers, Modules 1-3, 2007.
www.unesco.org/iiep
Wang, T. C., Lee, H. D., 2009. Developing a fuzzy
TOPSIS approach based on subjective weights and
objective weights, Expert Systems with Applications,
36(5), 8980–8985.
Wild, C., 1995. Continuous improvement of teaching: A
case study in a large statistics course, International
Statistical Review/Revue Internationale de Statistique,
63, 49-68.
Yang, J. L., Tzeng, G. H., 2011. An integrated MCDM
technique combined with DEMATEL for a novel
cluster-weighted with ANP method, Expert Systems
with Applications, 38(3), 1417-1424.
Yarmohammadian, M. H., Mozaffary, M., Esfahani, S. S.,
2011). Evaluation of quality of education in higher
education based on academic quality improvement
program (AQIP) model, Procedia Social and
Behavioral Sciences, 15, 2917–2922.
Yu, P. L., 1973. A class of solutions for group decision
problems, Management Science, 19(8), 936-946.
APPENDIX
A A HYBRID MCDM MODEL
COMBINED WITH DEMATEL
TECHNIQUE AND ANP
A.1 DEMATEL Technique
The DEMATEL technique is used to construct the
interactions/interrelationship between criteria to
build an influential relation map. The method is
divided into three steps:
Step 1: Find the average influence matrix
Α
The first step is to calculate initial matrix, using pair
of degree of interaction/interrelationship to obtain
directly influence matrix
Α
=
nnij
a
×
][ , where
ij
a
represents the degree of effect on i factor effects j
factor (Lin and Tzeng, 2009); (Chen et al., 2010).
Α
=
1
1
[] []
H
h
ij n n ij n n
h
aa
H
×
×
=
=
∑
(1)
where h is the h
th
expert and
1,2,...,hH=
.
Step 2: Calculate the normalized influence matrix
D
When the elements of i have a direct effect on the
elements of j, then
0≠
ij
a , otherwise 0=
ij
a . The
second step is to normalize the matrix. It can be
obtained from Eq. (2) and (3). Its diagonal is 0, and
maximum sum of row or column is 1.
s
=
D
A
(2)
,
11
11
min [1/ max ,1/ max ]
nn
ij ij
ij
in jn
ji
s
aa
≤≤ ≤ ≤
==
=
∑
∑
, 1,2,...,ij n=
(3)
Step 3: Compute the total influence matrix T
The total-influence matrix
T can be obtained through
Eq. (4), in which
I denotes the identity matrix.
21
... ( )
g −
=+ ++ = −TXX X XIX
when
lim [0]
g
nn
g
×
→∞
=X
(4)
Explanation:
T
2
g
=+ ++
X
XX
(
)
21 1
()()
h−−
=++++ −−XI X X X I X I X
1
()()
g −
=− −XI X I X , then
T
1
=( )
−
−XI X , when
lim [0]
g
g
nn→∞ ×
=X
.
STRATEGIESFORIMPROVINGACCREDITATIONPERFORMANCEINHIGHEREDUCATIONINSTITUTION
219

To sum of each row and column of the total effect
matrix
[]
ij n n
t
×
=T . Its will obtain the sum of all rows
(vector
11
1
1
[] (, , , , )
n
ij i n i n
j
n
trrrr
×
=
×
⎡⎤
′
===
⎢⎥
⎣⎦
∑
……r
) and
the sum of all columns (vector
1
1
1
[]
n
ij j n
i
n
td
×
=
×
⎡⎤
==
⎢⎥
⎣⎦
∑
d
1
(, , , , )
j
n
dd d= …… ). If
i
r
represents the sum of all
rows of the total-influence matrix
T, meaning
directly or/and indirectly affects to other criteria;
j
d represents the sum of all columns of the total-
influence matrix
T, meaning is affected by other
criteria.
i
r
represents the factor which will affect
other factors,
j
d represents the factor that is affected
by other factors. According to the definition,
ji
dr +
presents the degree of relationship between the
factors, meaning “prominence”;
ji
dr − presents the
degree of effect and effected for the factors, meaning
“relation” (Tzeng et al., 2007).
A.2 To find the Weights by DANP
Model
DANP is divided into following steps:
Step 1: Develop the structure of the question
The questions are clearly described then break them
down to level structure.
Step 2: Develop Unweighted Supermatrix
Firstly, each level with total degree of effect that
obtains from the total-influence matrix
T
of
DEMATEL as shown in Eq. (5).
1
1
11 1 1 1
1
11
12
1
1
1
2
1
2
11 1 1
1
1
cc
c
cc
c
c
c
cc
⎡⎤
⎢⎥
⎢⎥
⎢⎥
=
⎢⎥
⎢⎥
⎢⎥
⎢⎥
⎢⎥
⎣⎦














jn
mjjmnnm
n
j
i
n
c
c
c
m
c
i
c
i
c
im
i
c
n
c
n
c
nm
n
DD D
D
cc cc cc
jn
ij
iin
D
nnjnn
D
T
TTT
TTT
TTT
(5)
Normalize
c
T
with total-influence will be obtained
α
c
T
that shows in Eq. (6).
11
12
1
1
1
2
1
2
1
11 1 1 1
1
1
...
11 1 1
1
1
c
c
c
cc
ccc
jn
c
ij in
c
nnjnn
ααα
ααα
α
ααα
⎡⎤
⎢⎥
⎢⎥
⎢⎥
=
⎢⎥
⎢⎥
⎢⎥
⎢⎥
⎣⎦





 








c
c
c
m
ci
c
i
c
im
c
n
c
n
c
nm
n
jn
mjjm nnm
jn
i
i
n
DD D
cc cc cc
D
D
i
D
T
TTT
TTT
TTT
(6)
Normalize
11α
c
T
will be obtained by Eqs. (7) and (8),
according to the same fashion will be obtained
αnn
c
T
.
1
11
11
1
C
m
j
i
ij
dt
=
=
∑
,
1
1,2,...,im=
(7)
1
1
1
1
1111
11 1
11 11 11
11 11 11
11 1
11 1
11 11 11
11 11 11
11
1
11 11 11
11 11 11
///
// /
// /
m
CC
C
im
CC
C
C
mmjmm
CCC
j
ii i
iij
mm m
td td td
td td td
td td t d
α
⎡
⎤
⎢
⎥
⎢
⎥
⎢
⎥
⎢
⎥
=
⎢
⎥
⎢
⎥
⎢
⎥
⎢
⎥
⎢
⎥
⎣
⎦

 

 

T
=
1
1
1
1
11 11
11 11 11
11 1
11 11 11
1
11 11 11
m
CC
C
im
CC
C
mm mm
j
CC
C
j
iij
tt t
ttt
tt t
αα α
ααα
αα α
⎡
⎤
⎢
⎥
⎢
⎥
⎢
⎥
⎢
⎥
⎢
⎥
⎢
⎥
⎢
⎥
⎢
⎥
⎢
⎥
⎣
⎦

 

 

(8)
And then, total-influence matrix is normalized into
Supermatrix according to the group in relying
relationship to obtain Unweighted Supermatrix as
show in Eq. (9).
1
1
11 1 1 1
1
11
12
1
1
1
2
1
2
...
11 1 1
'
1
1
()
c
α
⎡
⎤
⎢
⎥
⎢
⎥
⎢
⎥
==
⎢
⎥
⎢
⎥
⎢
⎥
⎢
⎥
⎣
⎦














in
miimnnm
in
j
j
j
n
c
c
c
m
c
c
j
c
jm
c
n
c
n
c
nm
n
DD D
D
cc cc cc
in
D
jijnj
ninnn
D
T
WWW
W
WWW
WWW
(9)
In addition, we will be obtained matrix
11
W and
12
W by Eq. (10). If blank or 0 shown in the matrix
means the group or criteria is independent,
according to the same fashion will be obtained
matrix
nn
W
.
1
1
1
1
1111
11 1 1
11 11 11
11 1
1
11
11 11
11 11 11
1
1
1
11 11 11
1
()
m
cci
cm
cj cij
cm j
m
cm cim cmm
ttt
ttt
ttt
ααα
αα α
ααα
⎡
⎤
⎢
⎥
⎢
⎥
⎢
⎥
′
==
⎢
⎥
⎢
⎥
⎢
⎥
⎢
⎥
⎢
⎥
⎣
⎦



 


 
…
i
j
ccc
c
T
c
c
W
(10)
Step 3: Obtain Weight Supermatrix
Let each dimension of total-influence matrix
D
T
as
(11) be normalized with total degree of influence to
obtain
D
α
T
, the result as Eq. (12).
CSEDU2012-4thInternationalConferenceonComputerSupportedEducation
220

1
ij
n
j
iD
dt
=
=
∑
,
1,2,...,in=
1
11 1
1
1
=
DD
D
DD
D
DD
D
j
n
ij
iin
D
nj
nnn
ttt
ttt
ttt
⎡⎤
⎢⎥
⎢⎥
⎢⎥
⎢⎥
⎢⎥
⎢⎥
⎢⎥
⎢⎥
⎢⎥
⎣⎦

 

 

T
(11)

11 1 1
111
1
222
1
11 1 1
1
1
///
///
// /
DD
D
DD
D
DD
D
DD
D
DD
D
DD
D
jn
iijin
nnjnn
nn n
jn
iijin
nnjnn
D
td td td
td td td
td td td
ttt
tt t
tt t
ααα
αα α
αα α
α
=
⎡⎤
⎢⎥
⎢⎥
⎢⎥
=
⎢⎥
⎢⎥
⎢⎥
⎢⎥
⎢⎥
⎣⎦
⎡⎤
⎢⎥
⎢⎥
⎢⎥
⎢⎥
⎢⎥
⎢⎥
⎢⎥
⎢⎥
⎣⎦






 

 

T
(12)
Then, drive the normalized
D
α
T
into Unweight
Supermatrix
W
to obtain Weight Supermatrix
α
W
,
the result as shown in Eq. (13).
11 11 1 1 1 1
1
1
11
ii nn
DDD
jijnj
jij nj
D
DDD
nn inin nnnn
DDD
ttt
ttt
ttt
ααα
αα
ααα
ααα
⎡⎤
×××
⎢⎥
⎢⎥
⎢⎥
=×=
×××
⎢⎥
⎢⎥
⎢⎥
⎢⎥
×××
⎣⎦





WWW
WTW
WWW
WWW
(13)
Step 4: Obtain limit supermatrix
According to the weighted spuermatrix
α
W
, it
multiplies by itself multiple times to obtain limit
supermatrix. Then, the ANP weights of each
criterion can be obtained by
lim ( )
z
z
α
→∞
W
, where
z
represents any number for power.
B EVALUATING THE TOTAL
PERFORMANCE BY VIKOR
VIKOR can be divided into follow steps:
Step 1: Check the best value
*
j
f
and the worse
value
−
j
f
There
*
j
f
represents the positive-ideal point, that
means the expert gives the scores of the best value
(aspired levels) in each criterion and
−
j
f
represents
the negative-ideal point, that means the expert gives
the scores of the worst values in each criterion. We
use Eqs. (14) and (15) to obtain the results.
kjkj
ff max
*
=
,
nj ,,2,1 …
=
(traditional
approach)
or setting the aspired levels, vector
),,,(
**
2
*
1
*
n
ffff =
(14)
kjkj
ff min=
−
,
1,2,...,
j
n
=
(traditional
approach)
or setting the worst values,
vector
),,,(
21
−−−−
=
n
ffff 
(15)
Step 2: Calculate the mean of group utility
k
S
and
maximal regret
k
Q
.
There
k
S
represents the ratios of distance to the
positive-ideal, it means the synthesized gap for all
criteria;
j
w
represents the influential weights of the
criteria from DANP;
kj
r represents the average gap-
ratios (regret) of normalized distance to the aspired
level point, and
k
Q
represents the maximal gap-
ratios (regret) of normalized distance to the aspired
level in all criteria, it means the maximal gap in
j
criteria for prior improvement. Those values can
be computed respectively by Eqs. (16) and (17).
1
n
kjkj
j
Swr
=
=
∑
()()
1
n
j
jkj j j
j
wf f f f
∗
∗−
=
=−−
∑
(16)
{
}
njrQ
kjjk
,,2,1max …==
(17)
Step 3: Obtain the comprehensive indicator
k
R
and
sorting results.
The values can be computed respectively by Eq.
(18).
)/())(1()/()(
****
QQQQvSSSSvR
kk
−−−+−−=
−−
(18)
Those values derived from
kk
SS min
*
=
or setting
0
*
=S
(the aspired level),
kk
SS max=
−
or
setting
1=
−
S
(the worst situation);
kk
QQ min
*
=
or
setting
0
*
=Q (the aspired level), and
kk
QQ max=
−
or
setting
1=
−
Q (the worst situation). Therefore, when
0
*
=S
and
1=
−
S
, and 0
*
=Q and 1=
−
Q , we can re-
write the Eq.(18) as
kkk
QvvSR )1( −+=
. Weight
1
=
v
represents only to be consider the average gap
(average regret) weight and weight
0=v
represents
only to be consider the max gap to be prior
improvement. It can provide the decision-makers by
experts. Generally
5.0
=
v
(the majority of criteria), it
could be adjusted depends on the situation.
STRATEGIESFORIMPROVINGACCREDITATIONPERFORMANCEINHIGHEREDUCATIONINSTITUTION
221