Information Sciences 177 (2007) 3364–3382

www.elsevier.com/locate/ins

Using the analytic network process (ANP) in a SWOT

analysis – A case study for a textile firm
_
Ihsan Yüksel a, Metin Daǧdeviren b,*

a
Department of Business Management, Kirikkale University, 71450 Kirikkale, Turkey
b
Department of Industrial Engineering, Gazi University, Maltepe, 06570, Ankara, Turkey

Received 10 August 2005; received in revised form 29 December 2006; accepted 3 January 2007

Abstract

Strengths, Weaknesses, Opportunities and Threats (SWOT) analysis does not provide an analytical means to determine
the importance of the identified factors or the ability to assess decision alternatives according to these factors. Although the
analysis successfully pinpoints the factors, individual factors are usually described briefly and very generally. For this
reason, SWOT analysis possesses deficiencies in the measurement and evaluation steps. Although the analytic hierarchy
process (AHP) technique removes these deficiencies, it does not allow for measurement of the possible dependencies among
the factors. The AHP method assumes that the factors presented in the hierarchical structure are independent; however,
this assumption may be inappropriate in light of certain internal and external environmental effects. Therefore, it is
necessary to employ a form of SWOT analysis that measures and takes into account the possible dependency among
the factors. This paper demonstrates a process for quantitative SWOT analysis that can be performed even when there
is dependence among strategic factors. The proposed algorithm uses the analytic network process (ANP), which allows
measurement of the dependency among the strategic factors, as well as AHP, which is based on the independence between
the factors. Dependency among the SWOT factors is observed to effect the strategic and sub-factor weights, as well as to
change the strategy priorities.
Ó 2007 Elsevier Inc. All rights reserved.

Keywords: Strategic planning; SWOT analysis; Multi-criteria decision analysis; AHP; ANP

1. Introduction

Strategic management can be understood as the collection of decisions and actions taken by business man-
agement, in consultation with all levels within the organization, to determine the long-term activities of the
organization [8]. Many approaches and techniques can be used to analyze strategic cases in the strategic man-
agement process [2]. Among them, Strengths, Weaknesses, Opportunities and Threats (SWOT) analysis, which
evaluates the opportunities, threats, strengths and weaknesses of an organization, is the most common [7].

*
Corresponding author. Tel.: +903122317400; fax: +903122308434.
_ Yüksel), metindag@gazi.edu.tr (M. Daǧdeviren).
E-mail addresses: yuksel@kku.edu.tr (I.

0020-0255/$ - see front matter Ó 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.ins.2007.01.001
 _ Yüksel, M. Daǧdeviren / Information Sciences 177 (2007) 3364–3382
I. 3365

SWOT analysis is an important support tool for decision-making, and is commonly used as a means to sys-
tematically analyze an organization’s internal and external environments [11,13,15,36,40]. By identifying its
strengths, weaknesses, opportunities, and threats, the organization can build strategies upon its strengths,
eliminate its weaknesses, and exploit its opportunities or use them to counter the threats. The strengths
and weaknesses are identified by an internal environment appraisal while the opportunities and threats are
identified by an external environment appraisal [3]. SWOT analysis summarizes the most important internal
and external factors that may affect the organization’s future, which are referred to as strategic factors [11].
The external and internal environments consist of variables which are outside and inside the organization,
respectively. The organization’s management has no short-term effect on either type of variable [8].
Comprehensive environmental analysis is important in recognition of the variety of internal and external
forces with which an organization is confronted. On the one hand these forces may comprise potential stim-
ulants, and on the other hand, they may consist of potential limitations regarding the performance of the orga-
nization or the objectives that the organization wishes to achieve [8]. The obtained information can be
systematically represented in a matrix [38]; different combinations of the four factors from the matrix [2,8]
can aid in determination of strategies for long-term progress.
When used properly, SWOT can provide a good basis for strategy formulation [10]. However, SWOT anal-
ysis is not without weaknesses in the measurement and evaluation steps [7,20]. In conventional SWOT analysis,
the magnitude of the factors is not quantified to determine the effect of each factor on the proposed plan or strat-
egy [19]. In other words, SWOT analysis does not provide an analytical means to determine the relative impor-
tance of the factors, or the ability to assess the appropriateness of decision alternatives based on these factors
[10]. While it does pinpoint the factors in the analysis, individual factors are usually described briefly and very
generally [7]. More specifically, SWOT allows analysts to categorize factors as being internal (Strengths, Weak-
nesses) or external (Opportunities, Threats) in relation to a given decision, and thus enables them to compare
opportunities and threats with strengths and weaknesses [35]. However, the result of SWOT analysis is often
merely a listing or an incomplete qualitative examination of the internal and external factors [11]. For this rea-
son, SWOT analysis cannot comprehensively appraise the strategic decision-making process [7].

1.1. AHP in SWOT

Kurttila et al. [15] developed a hybrid method to eliminate the weaknesses in the measurement and evalu-
ation steps of the SWOT analysis [7,20]. This technique of utilizing the analytic hierarchy process (AHP) in the
SWOT analysis has been referred to as A’WOT in subsequent studies [10,18]. The hybrid method proposed by
Kurttila et al. [15] was tested in connection with a Finnish case study on forest certification. Examples in lit-
erature of studies that follow the method of Kurtilla et al. [15] include those by Stewart et al. [36], Kajanus
et al. [10], Shrestha et al. [35], Leskinen et al. [18] and Masozera et al. [19]. Similar to the work of Kurttila
et al. [15], these studies only deal with prioritization of the SWOT factors and sub-factors, and neither strat-
egies nor alternatives are included in the hierarchical structures based on the strategic factors.

1.2. Why ANP?

Although the AHP technique removes the deficiencies inherent in the measurement and evaluation steps of
SWOT analysis, it does not measure the possible dependencies among factors. The AHP method assumes that
the factors presented in the hierarchical structure are independent; however, this is not always a reasonable
presumption. The possible dependency among factors can only be determined as a result of internal and exter-
nal environmental analyses.
An organization can make good use of its opportunities if it possesses assets and capabilities in which it can
demonstrate superiority, otherwise opportunities are either lost before any benefit can be gained or are used by
rivals [38]. A similar relationship exists between threats and strengths. The ability to overcome or resist the
effects of threats depends on one’s strengths; a strong organization can use its strengths to either eliminate
or minimize the effects of these threats. The relationship between the weaknesses and strengths of an organi-
zation are such that an organization with more strengths would probably have fewer weaknesses, and there-
fore would be able to face situations arising from these weaknesses.
3366 _ Yüksel, M. Daǧdeviren / Information Sciences 177 (2007) 3364–3382
I.

Among the strategic factors, other two-variable combinations with possible interdependencies are threat-
weakness and opportunity-weakness. It can be claimed that organizations with more weaknesses than their
rivals are more susceptible to the threats. Thus, organizations should consider the relationship between their
threats and weaknesses when establishing their strategies. Similarly, an organization with weaknesses may find
it harder to make good use of its opportunities. It would be possible for an organization to benefit from the
opportunities if it has sufficient assets and capabilities [38], but if not, such opportunities arising from the
external environment may otherwise prove not useful [2].
As can be seen, the SWOT factors are not independent of each other, and moreover, there may even be a
relationship among some factors. Since the factor weights are traditionally computed by assuming that the
factors are independent, it is possible that the weights computed by including the dependent relations could
be different. Possible changes in the factor weights can change the priorities of alternative strategies, and these
changes, in turn, will affect the strategies chosen. Therefore, it is necessary to employ analyses which measure
and take the possible dependencies among factors into account in SWOT analysis.
In this study, SWOT analysis is performed using the analytical network process (ANP), which allows mea-
surement of dependency among SWOT factors. At the same time, the AHP method is used in order to deter-
mine the factor weights of the dependency or independency and their effects on the selection of an alternative
strategy. A comparative presentation of the results follows.

2. Analytic network process

An initial study identified the multi-criteria decision technique, known as the AHP, to be the most appro-
priate for solving complicated problems [17]. AHP was proposed by Saaty in 1980 [30] as a method of solving
socio-economic decision-making problems, and has been used to solve a wide range of decision-making prob-
lems [4,5,9,14].
AHP is a comprehensive framework which is designed to cope with the intuitive, the rational, and the irra-
tional when multi-objective, multi-criterion, and multi-actor decisions are made, with or without certainty, for
any number of alternatives. The basic assumption of AHP is the condition of functional independence of the
upper part, or cluster (see Fig. 1), of the hierarchy, from all its lower parts, and from the criteria or items in
each level [17].
Many decision-making problems cannot be structured hierarchically because they involve interaction of
various factors, with high-level factors occasionally depending on low-level factors [17,31,32]. Structuring a
problem with functional dependencies that allows for feedback among clusters is considered to be a network
system. Saaty suggested the use of AHP to solve the problem of independence among alternatives or criteria,
and the use of ANP to solve the problem of dependence among alternatives or criteria [17].
The ANP, also introduced by Saaty, is a generalization of the AHP [1,32]. While the AHP represents a
framework with a uni-directional hierarchical AHP relationship, the ANP allows for complex interrelation-
ships among decision levels and attributes. The ANP feedback approach replaces hierarchies with networks

cluster

••••

elements

Fig. 1. Structural difference between a hierarchy and a network [1]: (a) a hierarchy; (b) a network.
 _ Yüksel, M. Daǧdeviren / Information Sciences 177 (2007) 3364–3382
I. 3367

in which the relationships between levels are not easily represented as higher or lower, dominant or subordi-
nate, direct or indirect [22]. For instance, not only does the importance of the criteria determine the impor-
tance of the alternatives, as in a hierarchy, but the importance of the alternatives may also have an impact
on the importance of the criteria [32]. Therefore, a hierarchical representation with a linear top-to-bottom
structure is not suitable for a complex system [1].
A system with feedback can be represented by a network. The structural differences between a hierarchy
and a network are depicted in Fig. 1. The elements of a cluster may influence some or all the elements of
any other cluster. A network can be organized to include source clusters, intermediate clusters and sink clus-
ters. Relationships in a network are represented by arcs, where the directions of arcs signify directional depen-
dence [1,32]. Interdependency between two clusters, termed outer dependence, is represented by a two-way
arrow. Inner dependencies among the elements of a cluster are represented by looped arcs [1,33].
The ANP is composed of four major steps [1]:

Step 1: Model construction and problem structuring: The problem should be stated clearly and be decomposed
into a rational system, like a network. This network structure can be obtained by decision-makers
through brainstorming or other appropriate methods. An example of the format of a network is
shown in Fig. 1b.
Step 2: Pairwise comparison matrices and priority vectors: Similar to the comparisons performed in AHP, pairs
of decision elements at each cluster are compared with respect to their importance towards their control
criteria. The clusters themselves are also compared pairwise with respect to their contribution to the
objective. Decision-makers are asked to respond to a series of pairwise comparisons of two elements
or two clusters to be evaluated in terms of their contribution to their particular upper level criteria
[22]. In addition, interdependencies among elements of a cluster must also be examined pairwise;
the influence of each element on other elements can be represented by an eigenvector. The relative
importance values are determined with Saaty’s 1–9 scale (Table 1), where a score of 1 represents equal
importance between the two elements and a score of 9 indicates the extreme importance of one element
(row cluster in the matrix) compared to the other one (column cluster in the matrix) [22].A reciprocal
value is assigned to the inverse comparison, that is, aij ¼ 1=aji , where aij (aji ) denotes the importance
of the ith (jth) element. Like with AHP, pairwise comparison in ANP is performed in the framework
of a matrix, and a local priority vector can be derived as an estimate of the relative importance
associated with the elements (or clusters) being compared by solving the following equation:
A  w ¼ kmax  w ð1Þ
where A is the matrix of pairwise comparison, w is the eigenvector, and kmax is the largest eigenvalue of
A. Saaty [30] proposes several algorithms to approximate w. In this paper, Expert Choice [6] is used to
compute the eigenvectors from the pairwise comparison matrices and to determine the consistency
ratios.
Step 3: Supermatrix formation: The supermatrix concept is similar to the Markov chain process [32]. To
obtain global priorities in a system with interdependent influences, the local priority vectors are

Table 1
Saaty’s 1–9 scale for AHP preference [32]
Intensity of importance Definition Explanation
1 Equal importance Two activities contribute equally to the objective
3 Moderate importance Experience and judgement slightly favor one over another
5 Strong importance Experience and judgment strongly favor one over another
7 Very strong importance Activity is strongly favored and its dominance is demonstrated
in practice
9 Absolute importance Importance of one over another affirmed on the highest
possible order
2, 4, 6, 8 Intermediate values Used to represent compromise between the priorities listed above
Reciprocal of above If activity i has one of the above non-zero numbers assigned to it when compared with activity j, then j
non-zero numbers has the reciprocal value when compared with i
3368 _ Yüksel, M. Daǧdeviren / Information Sciences 177 (2007) 3364–3382
I.

entered in the appropriate columns of a matrix. As a result, a supermatrix is actually a partitioned

matrix, where each matrix segment represents a relationship between two clusters in a system. Let
the clusters of a decision system be C k ; k ¼ 1; 2; . . . ; n, and each cluster k has mk elements, denoted
by ek1 ; ek2 ; . . . ; ekmk . The local priority vectors obtained in Step 2 are grouped and placed in the appro-
priate positions in a supermatrix based on the flow of influence from one cluster to another, or from a
cluster to itself, as in the loop. A standard form for a supermatrix is as shown in expression (2) [32].
C1 Ck Cn
e 11 e 12 e1 m 1 e k1 ek2 ek mk en1 en2 en mn
e 11
C1 e 12

e1 m1 w 11 w 1k w 1n
e k1
e k2
ð2Þ
W = Ck w k1 w kk w kn
e k mk

e n1
Cn e n2 w n1 w nk w nn
en mn

As an example, the supermatrix representation for a hierarchy with three levels, as shown in Fig. 2a
[32], is as follows:
2 3
0 0 0
6 7
Wh ¼ 4 w21 0 0 5: ð3Þ
0 W32 I
In this matrix, w21 is a vector which represents the impact of the goal on the criteria, W32 is a matrix
that represents the impact of the criteria on each of the alternatives, I is the identity matrix, and zero
entries correspond to those elements having no influence.For the example given above, if the criteria
are interrelated, the hierarchy is replaced with the network shown in Fig. 2b. The interdependency is
exhibited by the presence of the matrix element W22 of the supermatrix Wn, yielding [32]:
2 3
0 0 0
6 7
W n ¼ 4 w21 W22 0 5: ð4Þ
0 W32 I
Note that any zero value in the supermatrix can be replaced by a matrix if there is an interrelationship
of the elements within a cluster or between two clusters. Since there usually is interdependence among
clusters in a network, the columns of a supermatrix may sum to more than one. However, the superm-
atrix must be modified so that each column of the matrix sums to unity. An approach recommended
by Saaty [32] involves determining the relative importance of the clusters in the supermatrix, using the

Goal Goal

w21 w21

Criteria Criteria
W22
W32
W32
Alternatives Alternatives

Fig. 2. Hierarchy and network [25]: (a) hierarchy; (b) network.

 _ Yüksel, M. Daǧdeviren / Information Sciences 177 (2007) 3364–3382
I. 3369

column cluster (see Fig. 1) as the controlling cluster. That is, row clusters with non-zero entries in a
given column cluster are compared according to their impact on the cluster of that column cluster. An
eigenvector is obtained from the pairwise comparison matrix of the row clusters with respect to the
column cluster, which in turn yields an eigenvector for each column cluster. The first entry of the
respective eigenvector for each column cluster, is multiplied by all the elements in the first cluster
of that column, the second by all the elements in the second cluster of that column and so on. In this
way, the cluster in each column of the supermatrix is weighted, and the result, known as the weighted
supermatrix, is stochastic.
Raising a matrix to exponential powers gives the long-term relative influences of the elements on each
other. To achieve convergence on the importance weights, the weighted supermatrix is raised to the
power of 2k þ 1, where k is an arbitrarily large number; the new matrix is called the limit supermatrix
[32]. The limit supermatrix has the same form as the weighted supermatrix, but all the columns of the
limit supermatrix are the same. The final priorities of all elements in the matrix can be obtained by
normalizing each cluster of this supermatrix. Additionally, the final priorities can be calculated using
matrix operations, especially where the number of elements in the model is relatively few. Matrix oper-
ations are used in order to easily convey the steps of the methodology and how the dependencies are
worked out.
Step 4: Selection of the best alternatives: If the supermatrix formed in Step 3 covers the whole network, the
priority weights of the alternatives can be found in the column of alternatives in the normalized
supermatrix. On the other hand, if a supermatrix only comprises clusters that are interrelated, addi-
tional calculations must be made to obtain the overall priorities of the alternatives. The alternative
with the largest overall priority should be selected, as it is the best alternative as determined by the
calculations made using matrix operations.

2.1. ANP literature

Over the years, ANP, a comprehensive multi-purpose decision method, has been widely used in solving
many complicated decision-making problems. In two studies by Meade and Sarkis [21,22], ANP was used
in a methodology they developed to evaluate logistic strategies and to improve production speed. Also, in
two separate studies performed by Lee and Kim [16,17], ANP is used in the interdependent information sys-
tem project selection process, and project priorities found in these two studies are taken as restraints in the 0–1
goal programming model. Karsak et al. [12] and Partovi and Corredoira [29] used ANP in a quality function
deployment process. In addition to these studies, other studies where ANP was used include: Meade and Pres-
ley [23], in evaluating alternative research-development projects; Sarkis [34], in a model he developed for the
purpose of strategic supplier selection; Mikhailov and Singh [24], in the development process of a decision sup-
port system; Yurdakul [41], in a model he built in order to evaluate long-term performances of production
systems; Momoh and Zhu [26], in specifying optimal production schedules; Niemira and Saaty [28], in Finan-
cial Crisis Forecasting; Ulutas [39], in determination of the appropriate energy policy for Turkey; and Chung
et al. [1], in a model they developed for a product mixture.

2.2. Proposed ANP algorithm for SWOT

The hierarchy and network model proposed in this study for SWOT analysis is composed of four levels, as
shown in Fig. 3. The goal (best strategy) is indicated in the first level, the criteria (SWOT factors) and sub-
criteria (SWOT sub-factors) are found in the second and third levels respectively, and the last level is com-
posed of the alternatives (alternative strategies).
The supermatrix of a SWOT hierarchy with four levels is as follows:
2 3
goal 0 0 0 0
SWOT factors 6 w21 0 0 07
W¼ 6 7; ð5Þ
SWOT sub-factors 4 0 W32 0 05
alternatives 0 0 W43 I
3370 _ Yüksel, M. Daǧdeviren / Information Sciences 177 (2007) 3364–3382
I.

Best Best
Goal Goal
strategy strategy
(w21) (w1)

SWOT SWOT (W2)

Criteria factors Criteria factors

(W32) (W3)

SWOT SWOT
Subcriteria sub-factors Subcriteria sub-factors

(W43) (W4)

Alternative Alternative
Alternatives Strategies Alternatives Strategies

Fig. 3. (a) The hierarchical representation of the SWOT model. (b) The network representation of the SWOT model.

where w21 is a vector which represents the impact of the goal on the criteria, W32 is a matrix that represents the
impact of the criteria on each of the sub-criteria, W43 is a matrix that represents the impact of the sub-criteria
on each of the alternatives, and I is the identity matrix.
A hierarchical representation of the SWOT model is given in Fig. 3a and its general network representation
is presented in Fig. 3b. The network model illustrates the case of a hierarchy with inner dependence within
clusters but no feedback. Here, SWOT factors, SWOT sub-factors and strategies are used in place of criteria,
sub-criteria and alternatives, respectively, and the SWOT factors have inner dependencies.
The main steps of our proposed framework can be summarized as follows. The first step of the study is the
identification of the SWOT factors, SWOT sub-factors and alternatives. The importance of the SWOT factor,
which corresponds to the first step of the matrix manipulation concept of the ANP, is determined based on the
works of Lee and Kim [17], and Saaty and Takizawa [31]. Then, according to the inner dependencies among
the SWOT factors, the inner dependency matrix, weights of SWOT sub-factors and priority vectors for alter-
native strategies based on the SWOT sub-factors are determined in given order.
The letters in parentheses in Fig. 3b represent the relationship that will be signified by sub-matrices for
supermatrix evaluation of the relative importance weights. Based on the schematic representation of
Fig. 3b, the general sub-matrix notation for the SWOT model used in this study is as follows:
2 3
goal 0 0 0 0
SWOT factors 6 w1 W2 0 0 7
W¼ 6 7; ð6Þ
SWOT sub-factors 4 0 W3 0 0 5
alternatives 0 0 W4 I
where w1 is a vector that represents the impact of the goal, namely, selecting the best strategy according to
SWOT factors, W2 is a matrix that represents the inner dependence of the SWOT factors, W3 is a matrix that
denotes the impact of the SWOT factor on each of the SWOT sub-factors, and W4 is a matrix that denotes the
impact of the SWOT sub-factors on each of the alternatives. Using matrix operations is preferred in order to
show the details of the calculations in this algorithm.
To apply the ANP to matrix operations in order to determine the overall priorities of the alternative strat-
egies identified with SWOT analysis, the proposed algorithm is as follows:

Step 1: Identify SWOT sub-factors and determine the alternative strategies according to SWOT sub-factors.
Step 2: Assume that there is no dependence among the SWOT factors; determine the importance degrees of
the SWOT factors with a 1–9 scale (i.e. calculate w1)
Step 3: Determine, with a 1–9 scale, the inner dependence matrix of each SWOT factor with respect to the
other factors by using the schematic representation of inner dependence among the SWOT factors:
(i.e. calculate W2)
 _ Yüksel, M. Daǧdeviren / Information Sciences 177 (2007) 3364–3382
I. 3371

Step 4: Determine the interdependent priorities of the SWOT factors (i.e. calculate wfactors ¼ W2  w1 )
Step 5: Determine the local importance degrees of the SWOT sub-factors with a 1–9 scale (i.e. calculate
wsub-factorsðlocalÞ )
Step 6: Determine the global importance degrees of the SWOT sub-factors (i.e. calculate wsub-factorsðglobalÞ ¼
wfactors  wsub-factorsðlocalÞ )
Step 7: Determine the importance degrees of the alternative strategies with respect to each SWOT sub-factor
with a 1–9 scale (i.e. calculate W4)
Step 8: Determine the overall priorities of the alternative strategies, reflecting the interrelationships within the
SWOT factors (i.e. calculate walternatives ¼ W4  wsub-factorsðglobalÞ )

3. An illustrative application of SWOT analysis

This section presents an illustration of the proposed approach summarized in the previous section. In the
following case study, SWOT analysis utilizing the ANP is performed on a firm which produces and exports
ready-made clothing in Istanbul, Turkey. The textile firm manufactures ready-made clothing from knitted fab-
ric and usually exports its products to European Union Countries.

3.1. Application of the proposed ANP model

In this study, first an external environment analysis is performed by an expert team familiar with the oper-
ation of the organization. In this way, those SWOT sub-factors which affect the success of the organization but
cannot be controlled by the organization are identified. In addition, an internal analysis is performed to deter-
mine the sub-factors which affect the success of the organization but can be controlled by the organization. In
based on these analyses, the strategically important sub-factors, i.e. the sub-factors which have very significant
effects on the success of the organization, are determined. Using the SWOT sub-factors, the SWOT matrix and
alternative strategies based on these sub-factors are developed (Table 2).
It can be seen from Table 2 that the organization has four alternative strategies. The strategy identified as
SO involves making good use of opportunities by using the existing strengths of the organization. The WO
strategy seeks to gain benefit from the opportunities presented by the external environmental factors by taking
into account the weaknesses of the organization. Similarly, ST is the strategy associated with using the
organization’s strengths to remove or reduce the effects of threats. The fourth and last strategy is WT, in which

Table 2
SWOT matrix
External factors Internal factors
Strengths (S) Weaknesses (W)
– Intellectual capital (S1) – Weak image of Turkish products (W1)
– Expert management staff (S2) – Energy costs (W2)
– Technically qualified workforce (S3) – Distance to market (W3)
– Quality of the product (S4)
Opportunities (O)
– Liberalization of Turkey (O1) SO Strategy WO Strategy
– New foreign markets (O2)
– Investment incentives (O3) Working with strong suppliers Making joint investments with EU suppliers
Threats (T)
– Threat of China (T1) ST Strategy WT Strategy
– Too high value of YTLa (T2)
– Economic and political uncertainty Investing in former east-bloc countries Subcontracting
in Turkey (T3)
– Current and possible problems in
Middle East (T4)
a
YTL: New Turkish Lira.
3372 _ Yüksel, M. Daǧdeviren / Information Sciences 177 (2007) 3364–3382
I.

the organization tries to reduce the effects of its threats by taking its weaknesses into account. In this study, the
aim of the SWOT analysis is to determine the priorities of the strategies developed and to determine the best
strategy for the organization.

Step 1: The problem is converted into a hierarchical structure in order to transform the sub-factors and alter-
native strategies into a state in which they can be measured by the ANP technique. The schematic
structure established is shown in Fig. 4. The aim of ‘‘choosing the best strategy’’ is placed in the first
level of the ANP model and the SWOT factors (Strengths, Weaknesses, Opportunities, Threats) are in
the second level. The SWOT sub-factors in the third level include: four sub-factors for the Strengths
factor, three sub-factors for the Weaknesses factor, three sub-factors for the Opportunities factor, and
three sub-factors for the Threats factor. Four alternative strategies developed for this study are placed
in the last level of the model. As presented in the SWOT matrix, these alternatives are as follows
(Table 2):
– Working with strong suppliers (SO);
– Making joint investments with EU suppliers (WO);
– Investing in former East-bloc countries (ST);
– Subcontracting (WT).
Step 2: Assuming that there is no dependence among the SWOT factors, pairwise comparison of the SWOT
factors using a 1–9 scale is made with respect to the goal. The comparison results are shown in Table 3.
All pairwise comparisons in the application are performed by the expert team mentioned in the begin-
ning of the study.
The pairwise comparison matrix, given in Table 3, is analyzed using Expert Choice software, and the
following eigenvector is obtained. In addition, the consistency ratio (CR) is provided in the last row of
the matrix.
2 3 2 3
S 0:447
6 W 7 6 0:282 7
6 7 6 7
w1 ¼ 6 7 ¼ 6 7
4 O 5 4 0:163 5
T 0:105

Intellectual Capital (S1)

Expert management staff (S2)
Technically Qualified workforce(S3) SO
Strengths Quality of the product (S4)
(S)
Weak image of Turkish products (W1)

Weaknesses Energy costs (W2) WO

Determining the (W)
Distance to market (W3)
best strategy
Oppurtunities Liberalization of Turkey (O1)
(O)
New Foreign Markets (O2) ST
Threats Investment incentives (O3)
(T)
Far East (T1)
Too high value of YTL (T2) WT
Economic and political uncertainty in
Turkey (T3)

Current and possible problems in

Middle East (T4)

Fig. 4. ANP model for SWOT.

 _ Yüksel, M. Daǧdeviren / Information Sciences 177 (2007) 3364–3382
I. 3373

Table 3
Pairwise comparison of SWOT factors by assuming that there is no dependence among them
SWOT factors S W O T Importance degrees of SWOT factors
Strengths (S) 1 2 3 3 0.447
Weaknesses (W) 1 2 3 0.282
Opportunities (O) 1 2 0.163
Threats (T) 1 0.105
CR = 0.03.

Step 3: Inner dependence among the SWOT factors is determined by analyzing the impact of each factor on
every other factor using pairwise comparisons. The introduction section mentioned that it is not
always possible to assume the SWOT factors to be independent. More appropriate and realistic results
can likely be obtained by using both SWOT analysis and the ANP technique. Using the analysis of
both the internal and external environments of the organization, the dependencies among the SWOT
factors, which are presented schematically in Fig. 5, are determined.
Based on the inner dependencies presented in Fig. 5, pairwise comparison matrices are formed for the
factors (Tables 4–6). The following question, ‘‘What is the relative importance of strengths when com-
pared with threats on controlling weaknesses?’’ may arise in pairwise comparisons and lead to a value of
9 (absolute importance) as denoted in Table 5. The resulting eigenvectors are presented in the last col-
umn of Tables 4–6.Using the computed relative importance weights, the inner dependence matrix of
the SWOT factors (W2) is formed. As opportunities are affected only by the Strengths, no pairwise
comparison matrix is formed for opportunities

O S W

Fig. 5. Inner dependence among SWOT factors.

Table 4
The inner dependence matrix of the SWOT factors with respect to ‘‘Strengths’’
Strengths W O T Relative importance weights
Weaknesses (W) 1 1/9 1/4 0.068
Opportunities (O) 1 3 0.681
Threats (T) 1 0.249
CR = 0.00.

Table 5
The inner dependence matrix of the SWOT factors with respect to ‘‘Weaknesses’’
Weaknesses S T Relative importance weights
Strengths (S) 1 9 0.900
Threats (T) 1 0.100
CR = 0.000.
3374 _ Yüksel, M. Daǧdeviren / Information Sciences 177 (2007) 3364–3382
I.

Table 6
The inner dependence matrix of the SWOT factors with respect to ‘‘Threats’’
Threats S W Relative importance weights
Strengths (S) 1 6 0.857
Weaknesses (W) 1 0.142
CR = 0.00.

2 3
1:000 0:900 1:000 0:857
6 0:068 1:000 0:000 0:142 7
6 7
W2 ¼ 6 7
4 0:681 0:000 1:000 0:000 5
0:249 0:100 0:000 1:000
Step 4: In this step, the interdependent priorities of the SWOT factors are calculated as follows:
2 3 2 3 2 3
1:000 0:900 1:000 0:857 0:447 0:478
6 0:068 1:000 0:000 0:142 7 6 7 6 7
6 7 6 0:282 7 6 0:164 7
wfactors ¼ W2  w1 ¼ 6 76 7¼6 7:
4 0:681 0:000 1:000 0:000 5 4 0:163 5 4 0:234 5
0:249 0:100 0:000 1:000 0:105 0:123
Significant differences are observed in the results obtained for the factor priorities (w1, Table 3) when
the interdependent priorities of the SWOT factors (wfactors ) and dependencies are ignored. The results
change from 0.447 to 0.478, 0.282 to 0.164, 0.163 to 0.234, and 0.105 to 0.123 for the priority values of
factors S, W, O and T, respectively.
Step 5: In this step, local priorities of the SWOT sub-factors are calculated using the pairwise comparison
matrix. The pairwise comparison matrices are detailed in Appendix A. Priority vectors obtained by
analyzing the pairwise comparison matrices provided in Appendix A are shown below.
2 3
0:411 2 3
6 0:300 7 0:625
6 7 6 7
wsub-factorsðstrengthsÞ ¼ 6 7 wsub-factorsðweaknessessÞ ¼ 4 0:238 5
4 0:098 5
0:136
0:180

Table 7
Overall priority of the SWOT sub-factors
SWOT Priority of SWOT sub-factors Priority of Overall priority of
factors the factors the sub-factors the sub-factors
Strengths 0.478 Intellectual capital (S1) 0.411 0.196
Expert management staff (S2) 0.300 0.143
Technically qualified workforce (S3) 0.098 0.047
Quality of the product (S4) 0.180 0.086
Weakness 0.164 Weak image of Turkish products (W1) 0.625 0.102
Energy costs (W2) 0.238 0.039
Distance to market (W3) 0.136 0.022
Opportunities 0.234 Liberalization of Turkey (O1) 0.178 0.041
New foreign markets (O2) 0.070 0.016
Investment incentives (O3) 0.751 0.176
Threats 0.123 Threat of China (T1) 0.282 0.034
Too high value of YTL (T2) 0.163 0.020
Economic and political uncertainty in Turkey (T3) 0.447 0.055
Current and possible problems in Middle East (T4) 0.105 0.013
 _ Yüksel, M. Daǧdeviren / Information Sciences 177 (2007) 3364–3382
I. 3375

2 3
2 3 0:282
0:178 6 0:163 7
6 7 6 7
wsub-factorsðoppurtunitiesÞ ¼ 4 0:070 5 wsub-factorsðthreatsÞ ¼ 6 7
4 0:447 5
0:751
0:105
Step 6: In this step, the overall priorities of the SWOT sub-factors are calculated by multiplying the interde-
pendent priorities of SWOT factors found in Step 4 with the local priorities of SWOT sub-factors
obtained in Step 5. The computations are provided in Table 7.The wsub-factorsðglobalÞ vector, obtained
by using the overall priority values of the sub-factors in the last column of Table 7, is provided below.
2 3
0:196
6 0:143 7
6 7
6 7
6 0:047 7
6 7
6 0:086 7
6 7
6 7
6 0:102 7
6 7
6 7
6 0:039 7
6 7
6 0:022 7
6 7
wsub-factorsðglobalÞ ¼ 6 7
6 0:041 7
6 7
6 0:016 7
6 7
6 7
6 0:176 7
6 7
6 7
6 0:034 7
6 7
6 0:020 7
6 7
6 7
4 0:055 5
0:013
Step 7: In this step we calculate the importance degrees of the alternative strategies with respect to each
SWOT sub-factors. The details of the pairwise comparison matrices are provided in Appendix B.
Using Expert Choice software, the eigenvectors are computed by analyzing these matrices and the
W4 matrix:
2 3
0:653 0:578 0:109 0:472 0:253 0:078 0:091 0:098 0:094 0:148 0:573 0:546 0:083 0:564
6 0:169 0:110 0:528 0:169 0:107 0:137 0:194 0:676 0:350 0:571 0:238 0:137 0:338 0:133 7
6 7
W4 ¼ 6 7:
4 0:111 0:240 0:051 0:284 0:573 0:469 0:577 0:161 0:508 0:226 0:131 0:232 0:535 0:242 5
0:065 0:071 0:310 0:072 0:065 0:314 0:137 0:064 0:047 0:053 0:056 0:083 0:042 0:060

Step 8: Finally, the overall priorities of the alternative strategies, reflecting the interrelationships within the
SWOT factors, are calculated as follows:
2 3 2 3
SO 0:366
6 WO 7 6 0:276 7
6 7 6 7
walternatives ¼ 6 7 ¼ W4  wsub-factorsðglobalÞ ¼ 6 7:
4 ST 5 4 0:271 5
WT 0:086
The ANP analysis results indicate that SO is the best strategy with an overall priority value of 0.366.

3.2. Comparing the AHP and ANP results

According to the ANP analysis, alternative strategies are ordered as SO–WO–ST–WT. The same example is
analyzed with the hierarchical model given in Fig. 3a by assuming there is no dependence among the factors.
3376 _ Yüksel, M. Daǧdeviren / Information Sciences 177 (2007) 3364–3382
I.

Table 8
Weights and ranking of the strategies with AHP and ANP
SO WO ST WT
Weights in AHP 0.356 0.240 0.308 0.094
Ranking in AHP 1 3 2 4
Weights in ANP 0.366 0.276 0.271 0.086
Ranking in ANP 1 2 3 4

The overall priorities computed for the alternative strategies are presented below. The same pairwise compar-
ison matrices are used to compute the AHP priority values.
2 3 2 3
SO 0:356
6 WO 7 6 0:240 7
6 7 6 7
walternativesðAHPÞ ¼ 6 7¼6 7
4 ST 5 4 0:308 5
WT 0:094
In the AHP analysis, the SO strategy is found to be the best alternative, with an overall priority value of
0.356. However, the priority ordering of the alternative strategies is changed to SO–ST–WO–WT. When
dependence among factors is taken into account, both the strategy priorities and the ranking order of the strat-
egies changes. The results obtained from the AHP and ANP analyses are comparatively listed in Table 8.
In cases where the dependency among SWOT factors and sub-factors is established, ANP analysis can be
performed in order to determine the alternative priorities so that firms are able to make strategically correct
decisions. AHP analysis can be used in situations where there is no dependency among SWOT factors and sub-
factors or where the level of this dependency can be neglected.

3.3. Validation of the model

The importance of testing the validity of models that are developed in operation research studies is a well-
known fact. However, it can be seen that the validity of the theoretical base [32] of the suggested ANP model
has been neglected and specific criteria for testing it have not been developed. This shortcoming can be seen in
any study that applies the ANP technique [16,17,21,22,28,39,41]. In addition, there are only two practical
studies [27,37] that use the AHP technique in literature and debate the validity of the suggested model. Similar
to other works, the present study has also faced some limitations and difficulties with testing the validity of the
suggested model.
The first of these difficulties stems from the fact that the factors in the ANP model are not quantitative by
nature. ANP is a technique that is used in solving multiple-criteria decision-making problems where there is
dependency between factors that are both qualitative and quantitative in nature. Moreover, problems that are
modeled by the ANP pairwise comparison matrices used to determine the priority values for the factors are
determined by the judgment of experts. However, it is not always possible to objectively assign numerical mea-
surements to the elements in a decision-making problem, nor is it possible to come up with the same results
each time. This is because the data used in pairwise comparison matrices may change depending on the sub-
jective views of the experts. Thus it is impossible to arrive at the same results using data obtained from dif-
ferent case studies. However, this limitation is embedded in the very nature of decision-making problems.
Likewise, it is a well-known fact that different preferences are made under different conditions. Consequently,
the fact that the values of the pairwise comparative factors change depending on the views of experts should
not be a reason for rejecting the validity of studies using the suggested ANP model.
Another problem that is encountered when testing the validity of the model is that the model has not been
analyzed using past data, due to the unavailability of past data for the particular management case under
study. This problem, however, should not be viewed as a significant shortcoming when evaluating the validity
of the model. The comparison matrices that are the inputs to the suggested model are defined under known
conditions. Thus it is possible to achieve different results since different pairwise comparison matrices may be
obtained at different points in time.
 _ Yüksel, M. Daǧdeviren / Information Sciences 177 (2007) 3364–3382
I. 3377

In consideration of the problems mentioned above, the validity of the proposed model in this study was
debated and evaluated in two ways. First, the results from the proposed model were compared with AHP
model. The results obtained using the proposed model are presented in Section 3.1, and demonstrate the
best-choice strategy for the case where there is dependency between the SWOT factors. Different results are
obtained when the ANP and AHP methods are implemented, despite them using the same pairwise compar-
ison matrices (Table 8). However, such a difference is expected because AHP does not take into account
dependencies among factors while ANP does. For this reason, the ANP method is better able to model real
world situations as compared to the AHP method. The superiority of ANP allows SWOT analysis to yield
more realistic results. Not only has the suggested model enabled us to satisfy the objective of our study but
it has also demonstrated he functionality of the model.
The second criterion for model validity in this study was the views of management strategists. This criterion
was used in the AHP technique in this study. The management strategists of the firm on which the case study
was conducted have said that they have found the results obtained from the suggested model to be meaningful
and useful. Thus, as of August 2006, in view of the strategies followed in the last year the strategists claim that
the strategies determined by the model in our study in May 2005 are indeed accepted as the ‘‘best solutions.’’
Even though the suggested model was applied to a textile firm, its structure can be modified and used for
other industries. There are two reasons why the model may be in need of modification. The first reason is that
the sub-factors and alternative strategies that the model incorporates may need to be changed to suit manage-
ment. Each management team must incorporate strategies into the model based on the strategic factors which
it has determined using the results of its internal and external analyses. The second reason is that the depen-
dency between SWOT factors and sub-factors may vary depending on the type of management. For example,
only the dependency between SWOT factors was found to be important for the firm under study in this
research. It may be that, for a different management factor, the dependency between the factors that form
the sub-factors is important in addition to the dependency between the SWOT factors. The model suggested
in the study is able to maintain its utility in the face of different types of management.
Another parameter that verifies the validity of the model is the consistency ratio (CR) of the pairwise com-
parison matrices. The consistency ratio in the pairwise comparison matrices is calculated using the consistency
index and the random index [30]. The consistency index (CI) of a matrix of comparisons is given by
CI ¼ ðkmax  nÞ=ðn  1Þ. Here, kmax is the maximum eigenvalue and n is the size of matrix. The consistency
ratio (CR) is obtained by forming the ratio of the CI and the random index (RI). DeSchutter has conjectured
the following relationship between the RI and n: RI ¼ 1:98  ½ðn  2Þ=n, where 1.98 is the average value of
the ratio of each value computed so far from n ¼ 3 to 15 multiplication by ðn  2Þ=n for the corresponding
value of n [5]. The value of the CR calculated using AHP and ANP must be less than 0.10. The consistency
ratio of the pairwise comparison matrices used in this study were obtained using Expert Choice [6], and the
relevant values can be found in the last line of the pairwise comparison matrices given in Appendices A
and B. Upon examination of all the calculated consistency ratios, it can be seen that all the values are less than
0.10. Given these consistency ratios, we may be confident of the appropriateness of the pairwise comparison
matrices used in this study.

4. Conclusions

In SWOT analysis, strategic alternatives are selected in the light of the strengths, weaknesses, threats and,
opportunities of the organization as determined through internal and external environment analysis. How-
ever, SWOT analysis is not capable of quantitatively determining the weights and effects of the strategic fac-
tors on the alternatives. Although some studies do perform such quantitative weighting, these studies fail to
consider the relations or dependencies of the factors of the SWOT analysis. It is generally not possible to
assume the SWOT factors to be independent and unrelated with one another.
In this study, we sought to demonstrate, with a case study example, that it is possible to perform a quan-
titative SWOT analysis wherein the possible dependencies among SWOT factors are included. The ANP tech-
nique, which enables measuring inter-factor dependencies, is utilized in this work. The AHP method is also
used with SWOT analysis to compare the effects of the dependency among the SWOT factors on prioritizing
the alternative strategies and on the SWOT sub-factors weights.
3378 _ Yüksel, M. Daǧdeviren / Information Sciences 177 (2007) 3364–3382
I.

The factors from the SWOT analysis and the alternative strategies based on these factors were transformed
into an ANP model. The first four levels of the ANP model consist of the aim (choosing the best strategy), the
SWOT factors, SWOT sub-factors and, four alternative strategies, respectively. The weights of the factors of
the SWOT analysis differ according to the method used in the analysis (AHP or ANP), due to the dependency
among the SWOT factors. In other words, the dependency or independency of SWOT factors affects the factor
weights of the SWOT analysis. In this study, it was observed that the dependency among SWOT factors affects
both the strategy selection and the strategy priority order. The strategy orders obtained by the ANP method
and the AHP method were found to be different.
Future research may seek to also consider the effect of possible dependencies among the SWOT sub-factors
themselves. In addition, fuzzy numbers can be introduced in the AHP or ANP methods to more effectively
analyze cases having greater uncertainty in the pairwise comparison matrices.

Acknowledgements

The authors gratefully acknowledge the Editor-in-Chief, Prof. Witold Pedrycz, and reviewers’ valuable sug-
gestions and constructive comments on the earlier version of this paper.

Appendix A. Pairwise comparison matrices for SWOT sub-factors local priorities

Strengths S1 S2 S3 S4 Local weights

Intellectual capital (S1) 1 2 3 2 0.411
Expert management staff (S2) 1 4 2 0.309
Technically qualified workforce (S3) 1 1/2 0.098
Quality of the product (S4) 1 0.180
CR = 0.04.

Weaknesses W1 W2 W3 Local weights

Weak image of Turkish products (W1) 1 3 4 0.625
Energy costs (W2) 1 2 0.238
Distance to market (W3) 1 0.136
CR = 0.02.

Opportunities O1 O2 03 Local weights

Liberalization of Turkey (O1) 1 3 1/5 0.178
New foreign markets (O2) 1 1/9 0.070
Investment incentives (O3) 1 0.751
CR = 0.03.

Threats T1 T2 T3 T4 Local weights

Threat of China (T1) 1 2 1/2 3 0.282
High value of YTL (T2) 1 1/3 2 0.163
Economic and political uncertainty in Turkey (T3) 1 3 0.447
Current and possible problems in the Middle East (T4) 1 0.105
CR = 0.03.
 _ Yüksel, M. Daǧdeviren / Information Sciences 177 (2007) 3364–3382
I. 3379

Appendix B. Pairwise comparison matrices for the priorities of the alternative strategies based on the SWOT sub-
factors

Intellectual capital (S1) SO WO ST WT Local weights

SO Working with strong suppliers 1 6 5 7 0.653
WO Making joint investments with EU suppliers 1 2 3 0.169
ST Investing in former East-bloc countries 1 2 0.111
WT Subcontracting 1 0.065
CR = 0.04.

Expert management staff (S2) SO WO ST WT Local weights

SO Working with strong suppliers 1 5 3 7 0.578
WO Making joint investments with EU suppliers 1 1/3 2 0.110
ST Investing in former East-bloc countries 1 3 0.240
WT Subcontracting 1 0.071
CR = 0.02.

Technically qualified workforce (S3) SO WO ST WT Local weights

SO Working with strong suppliers 1 1/4 3 1/4 0.109
WO Making joint investments with EU suppliers 1 9 2 0.528
ST Investing in former East-bloc countries 1 1/5 0.051
WT Subcontracting 1 0.310
CR = 0.03.

Quality of the product (S4) SO WO ST WT Local weights

SO Working with strong suppliers 1 3 2 5 0.472
WO Making joint investments with EU suppliers 1 1/2 3 0.169
ST Investing in former East-bloc countries 1 4 0.284
WT Subcontracting 1 0.072
CR = 0.02.

Weak image of Turkish products (W1) SO WO ST WT Local weights

SO Working with strong suppliers 1 3 1/3 4 0.253
WO Making joint investments with EU suppliers 1 1/5 2 0.107
ST Investing in former East-bloc countries 1 7 0.573
WT Subcontracting 1 0.065
CR = 0.02.

Energy costs (W2) SO WO ST WT Local weights

SO Working with strong suppliers 1 1/2 1/5 1/4 0.078
WO Making joint investments with EU suppliers 1 1/3 1/3 0.137
ST Investing in former East-bloc countries 1 2 0.469
WT Subcontracting 1 0.314
CR = 0.02.
3380 _ Yüksel, M. Daǧdeviren / Information Sciences 177 (2007) 3364–3382
I.

Distance to market (W3) SO WO ST WT Local weights

SO Working with strong suppliers 1 1/2 1/5 1/2 0.091
WO Making joint investments with EU suppliers 1 1/4 2 0.194
ST Investing in former East-bloc countries 1 4 0.577
WT Subcontracting 1 0.137
CR = 0.03.

Liberalization of Turkey (O1) SO WO ST WT Local weights

SO Working with strong suppliers 1 1/7 1/2 2 0.098
WO Making joint investments with EU suppliers 1 6 7 0.676
ST Investing in former East-bloc countries 1 3 0.161
WT Subcontracting 1 0.064
CR = 0.04.

New foreign markets (O2) SO WO ST WT Local weights

SO Working with strong suppliers 1 1/5 1/6 3 0.094
WO Making joint investments with EU suppliers 1 1/2 8 0.350
ST Investing in former East-bloc countries 1 7 0.508
WT Subcontracting 1 0.047
CR = 0.05.

Investment incentives (O3) SO WO ST WT Local weights

SO Working with strong suppliers 1 1/3 1/2 3 0.148
WO Making joint investments with EU suppliers 1 0.571
ST Investing in former East-bloc countries 1 0.226
WT Subcontracting 1 0.053
CR = 0.04.

Threat of China (T1) SO WO ST WT Local weights

SO Working with strong suppliers 1 3 5 7 0.573
WO Making joint investments with EU suppliers 1 2 5 0.238
ST Investing in former East-bloc countries 1 3 0.131
WT Subcontracting 1 0.056
CR = 0.03.

High value of YTL (T2) SO WO ST WT Local weights

SO Working with strong suppliers 1 4 3 5 0.546
WO Makiqng joint investments with EU suppliers 1 1/2 2 0.137
ST Investing in former East-bloc countries 1 3 0.232
WT Subcontracting 1 0.083
CR = 0.02.
 _ Yüksel, M. Daǧdeviren / Information Sciences 177 (2007) 3364–3382
I. 3381

Economic and political uncertainty in Turkey (T3) SO WO ST WT Local weights

SO Working with strong suppliers 1 1/5 1/8 3 0.083
WO Making joint investments with EU suppliers 1 1/2 9 0.338
ST Investing in former East-bloc countries 1 8 0.535
WT Subcontracting 1 0.042
CR = 0.05.

Current and possible problems in the Middle East (T4) SO WO ST WT Local weights
SO Working with strong suppliers 1 5 3 6 0.564
WO Making joint investments with EU suppliers 1 1/2 3 0.133
ST Investing in former East-bloc countries 1 5 0.242
WT Subcontracting 1 0.060
CR = 0.04.

References

[1] S.H. Chung, A.H.L. Lee, W.L. Pearn, Analytic network process (ANP) approach for product mix planning in semiconductor
fabricator, International Journal of Production Economics 96 (2005) 15–36.
[2] O. Dincer, Strategy Management and Organization Policy, Beta Publication, Istanbul, 2004.
[3] R.G. Dyson, Strategic development and SWOT analysis at the University of Warwick, European Journal of Operational Research
152 (2004) 631–640.
[4] Y.C. Erensal, T. Özcan, M.L. Demircan, Determining key capabilities in technology management using fuzzy analytic hierarchy
process: a case study of Turkey, Information Sciences 176 (2006) 2755–2770.
[5] T. Ertay, D. Ruan, U.R. Tuzkaya, Integrating data envelopment analysis and analytic hierarchy for the facility design in
manufacturing systems, Information Sciences 176 (2006) 237–262.
[6] Expert Choice, Expert Choice, Analytical Hierarchy Process (AHP) Software, Version 9.5, Expert Choice, Pittsburg, 2000.
[7] T. Hill, R. Westbrook, SWOT analysis: it’s time for a product recall, Long Range Planning 30 (1997) 46–52.
[8] G. Houben, K. Lenie, K. Vanhoof, A knowledge-based SWOT-analysis system as an instrument for strategic planning in small and
medium sized enterprises, Decision Support Systems 26 (1999) 125–135.
[9] C. Kahraman, D. Ruan, I. _ Doğan, Fuzzy group decision-making for facility location selection, Information Sciences 157 (2003) 135–153.
[10] M. Kajanus, J. Kangas, M. Kurttila, The use of value focused thinking and the A’WOT hybrid method in tourism management,
Tourism Management 25 (2004) 499–506.
[11] J. Kangas, M. Kurtila, M. Kajanus, A. Kangas, Evaluating the management strategies of a forestland estate-the S-O-S approach,
Journal of Environmental Management 69 (2003) 349–358.
[12] E.E. Karsak, S. Sozer, S.E. Alptekin, Production planning in quality function deployment using a combined analytical network
process and goal programming approach, Computers and Industrial Engineering 44 (2002) 171–190.
[13] P. Kotler, Marketing Management: Analysis, Planning, Implementation and Control, Prentice-Hall, New Jersey, 1988.
[14] O. Kulak, C. Kahraman, Fuzzy multi-attribute selection among transportation companies using axiomatic design and analytic
hierarchy process, Information Sciences 170 (2005) 191–210.
[15] M. Kurttila, M. Pesonen, J. Kangas, M. Kajanus, Utilizing the analytic hierarchy process (AHP) in SWOT analysis-a hybrid method
and its application to a forest-certification case, Forest Policy and Economics 1 (2000) 41–52.
[16] J.W. Lee, S.H. Kim, An integrated approach for independent information system project selection, International Journal of Project
Management 19 (2001) 111–118.
[17] J.W. Lee, S.H. Kim, Using analytic network process and goal programming for interdependent information system project selection,
Computers and Operations Research 27 (2000) 367–382.
[18] L.A. Leskinen, P. Leskinen, M. Kurttila, J. Kangas, M. Kajanus, Adapting modern strategic decision support tools in the
participatory strategy process-a case study of a forest research station, Forest Policy and Economics 8 (2006) 267–278.
[19] M.K. Masozera, J.R.R. Alavalapati, S.K. Jacobson, R.K. Shresta, Assessing the suitability of community-based management for the
Nyungwe Forest Reserve, Rwanda, Forest Policy and Economics 8 (2006) 206–216.
[20] M.H.B. McDonald, The Marketing Planner, Butter-worth-Heinemann, Oxford, 1993.
[21] L.M. Meade, J. Sarkis, Strategic analysis of logistics and supply chain management systems using the analytical network process,
Transportation Research Part E: Logistics and Transportation Review 34 (1998) 201–215.
[22] L.M. Meade, J. Sarkis, Analyzing organizational project alternatives for agile manufacturing processes: an analytical network
approach, International Journal of Production Research 37 (1999) 241–261.
[23] L.M. Meade, A. Presley, R&D project selection using the analytic network process, IEEE Transactions on Engineering Management
49 (2002) 59–66.
3382 _ Yüksel, M. Daǧdeviren / Information Sciences 177 (2007) 3364–3382
I.

[24] L. Mikhailov, M.S. Singh, Fuzzy analytic network process and its application to the development of decision support systems, IEEE
Transactions on Systems, Man, and Cybernetics-Part C: Applications and Reviews 33 (2003) 33–41.
[25] J.A. Momoh, J.Z. Zhu, Application of AHP/ANP to unit commitment in the deregulated power industry, In: 1998 IEEE
International Conference on Systems, Man and Cybernetics, vol. 1 San Diego, 1998, pp. 817–822.
[26] J.A. Momoh, J. Zhu, Optimal generation-scheduling based on AHP/ANP, IEEE Transactions on Systems, Man, and Cybernetics-
Part B: Cybernetics 33 (2003) 531–535.
[27] E.W.T. Ngai, Selection of web sites for online advertising using the AHP, Information and Management 40 (2003) 233–242.
[28] M.P. Niemira, T.L. Saaty, An analytical network process model for financial-crisis forecasting, International Journal of Forecasting
20 (2004) 573–587.
[29] F.Y. Partovi, R.A. Corredoira, Quality function deployment for the good of soccer, European Journal of Operational Research 137
(2002) 642–656.
[30] T.L. Saaty, The Analytic Hierarchy Process, McGraw-Hill, New York, 1980.
[31] T.L. Saaty, M. Takizawa, Dependence and independence: from linear hierarchies to nonlinear Networks, European Journal of
Operational Research 26 (1986) 229–237.
[32] T.L. Saaty, Decision Making with Dependence and Feedback: The Analytic Network Process, RWS Publications, Pittsburgh, 1996.
[33] J. Sarkis, Quantitative models for performance measurement systems-alternate considerations, International Journal of Production
Economics 86 (2002) 81–90.
[34] J. Sarkis, A model for strategic supplier selection, Journal of Supply Chain Management 38 (2002) 18–28.
[35] R.K. Shrestha, J.R.R. Alavalapati, R.S. Kalmbacher, Exploring the potential for silvopasture adoption in South-central Florida: an
application of SWOT-AHP method, Agricultural Systems 81 (2004) 185–199.
[36] R. Stewart, S. Moamed, R. Daet, Strategic implementation of IT/IS projects in construction: a case study, Automation in
Construction 11 (2002) 681–694.
[37] D. Thirumalaivasan, M. Karmegam, K. Venugopal, AHP-DRASTIC: software for specific aquifer vulnerability assessment using
DRASTIC model and GIS, Environmental Modelling and Software 18 (2003) 645–656.
[38] H. Ulgen, S.K. Mirze, Strategic Management, Literatur Publication, Istanbul, 2004.
[39] B.H. Ulutasß, Determination of the appropriate energy policy for Turkey, Energy 30 (2005) 1146–1161.
[40] T.L. Wheelen, J.D. Hunger, Strategic Management and Business Policy, Addison-Wesley, Reading, MA, 1995.
[41] M. Yurdakul, Measuring long-term performance of a manufacturing firm using the analytical network process (ANP) approach,
International Journal of Production Research 41 (2003) 2501–2529.