TECHNOLOGICAL AND ECONOMIC DEVELOPMENT OF ECONOMY

ISSN 2029-4913 / eISSN 2029-4921

2016 Volume 22(2): 210–234

doi:10.3846/20294913.2014.984363

FUZZY MULTI-CRITERIA DECISION MAKING MODEL

FOR SUBCONTRACTOR SELECTION IN INTERNATIONAL
CONSTRUCTION PROJECTS

Serdar ULUBEYLIa, Aynur KAZAZb

aDepartment of Civil Engineering, Faculty of Engineering, Bulent Ecevit University,
67100 Zonguldak, Turkey
bDepartment of Civil Engineering, Faculty of Engineering, Akdeniz University,

07058 Antalya, Turkey

Received 18 October 2012; accepted 25 August 2013

Abstract. A general contractor’s ability to select proper subcontractors in foreign projects is a key
competitive advantage. Toward this aim, a subcontractor selection model (CoSMo) was developed in
this study. As a computational approach, the fuzzy sets method was employed because it can model
human judgment by means of linguistic values, combining qualitative and quantitative decision
criteria into an aggregate measure. Although the algorithm may be complex for easy acceptance by
industrial practitioners, this disadvantage was minimized through a computer-supported system.
In order to gain a better understanding of the current practice of CoSMo, a real world construction
project was conducted. As a result, it was observed that CoSMo has high practical application and
can be used as an advisory system by satisfying principal contractor’s requirements to reduce the risk
involved in the selection of a subcontractor. Moreover, it gives an initial idea of how subcontractors
perform on each decision criterion and allows the main contractor to understand the picture on
the strong and weak points of each bidder and thereby to take conscious decisions.

Keywords: subcontractor selection, fuzzy sets, multiple criteria decision making, international
construction projects.

JEL Classification: D81, L24, L74, M15, M16, M51.

Introduction

The subcontracting system is usually described as the contractual process in which a principal
contractor subcontracts some parts of the work to another contractor (Chiang 2009; Lingard
et al. 2010). In the construction industry, many general contractors act as project agents,
and transfer actual project tasks to subcontractors for execution (Shash 1998; Grasso et al.

Corresponding author Serdar Ulubeylı

E-mail: [email protected]

Copyright © 2016 Vilnius Gediminas Technical University (VGTU) Press

http://www.tandfonline.com/TTED
Technological and Economic Development of Economy, 2016, 22(2): 210–234 211

2008; Ng, Tang 2010). Subcontractors may supply workers, materials, equipment, tools, and
even design (Tam et al. 2011). They could bring unique skills and talents for specialized
work such as steel work, concreting, installation, and other sophisticated facility systems
(Hinze, Tracey 1994). Since a huge proportion of work is conducted by subcontractors, the
non-performance of any subcontracting organization is one of the main causes for project
failure (Arditi, Chotibhongs 2005; De Silva et al. 2012).
Improvements in subcontractor selection processes have not received the attention that
one would expect from a significant contribution to the construction industry. A construc-
tion project which lacks a rigorous subcontractor selection process is prone to many serious
problems such as time and cost overruns, substandard works, disputes, and finally dissatis-
faction of clients. From another perspective, it may also lead to the resignation of qualified
subcontractors from the business or result in a lowering of their standards; thus producing
cheap and poor-quality work (Choudhry et al. 2012; Kazaz et al. 2012).

1. Previous researches

Although numerous researches on the prequalification process of main contractors have been
carried out to date (Nguyen 1985; Juang et al. 1987; Paek et al. 1992; Elton et al. 1994; Singh,
Tiong 2005; Brauers et al. 2008; Mitkus, Trinkuniene 2008; Turskis 2008; Zavadskas et al. 2008,
2009, 2010; Plebankiewicz 2009, 2010; Podvezko et al. 2010; Arslan 2012; Cheng, Kang 2012;
San Cristobal 2012; Vahdani et al. 2013; Wang et al. 2013), only a number of scientific papers
discussing the problem of construction subcontractor prequalification have been published.
Ng and Luu (2008) developed a case-based system for construction subcontractor registra-
tion. Mbachu (2008) suggested a framework for use by main contractors and consultants
in the assessment of the suitability and performance of subcontractors. Arslan et al. (2008)
proposed a web-based subcontractor evaluation system. Ip et al. (2004) made a specific re-
search on subcontractor selection in a construction project, employing a mathematical model
developed for a specific use in the manufacturing industry and based on cost and duration.
Tserng and Lin (2002) developed an Internet-based model in order to take proposals from
subcontractors. They take into account risk and profitability as well as reciprocal variations
between these two factors by means of cost and time schedules. Kumaraswamy and Matthews
(2000) suggested a system based on partnering relationships for the choice of subcontractors.
They investigated the effect level of project-based or long-term partnership on the selection.
In their study, a general framework was created without developing a specific model. Okoroh
and Torrance (1999) proposed a model for subcontractor selection in refurbishment projects.
They have presented a model for analyzing the subcontractor’s risk elements in construction
refurbishment projects and described a prototype knowledge-based expert system. Albino
and Garavelli (1998) used neural networks for the selection of subcontractors. However, these
studies have some drawbacks which can be listed in the following items:
–– Ip et al. (2004), Tserng and Lin (2002), Okoroh and Torrance (1999), and Albino and
Garavelli (1998) considered very few selection criteria and formed one-tier selection
procedure, which made the selection process superficial,
212 S. Ulubeyli, A. Kazaz. Fuzzy multi-criteria decision making model for subcontractor ...

–– Ip et al. (2004) and Tserng and Lin (2002) did not take into account qualitative criteria,
which was obviously in contradiction with the characteristics of a real life problem,
–– Ng and Luu (2008), Okoroh and Torrance (1999), and Albino and Garavelli (1998)
depended solely on past data that should be collected and trained to solve the problem,
which produced a time-consuming model in the short- and mid-term,
–– Mbachu (2008) and Ip et al. (2004) did not employ any automated system, which made
the calculation procedure an effort-consuming process,
–– Okoroh and Torrance (1999) has project-specific criteria, which did not allow the
model to be easily used for all kinds of projects, and
–– Arslan et al. (2008) evaluated criteria that were scored on a 1 to 10 scale, which was,
in fact, not adequately suitable for decision makers because of the fact that human
perception and judgment cannot be quantified precisely and that decision makers
intuitively feel more comfortable providing their judgments in verbal terms (rather
than numerically), which, due to subjectivity, leads to ambiguity in human decision
making (Poyhonen et al. 1997).

2. Objectives of the study

This study intends to develop a construction subcontractor selection model, called CoSMo,
by taking into account the above-mentioned shortcomings of previous researches.
Although CoSMo can be employed in domestic construction projects, it was developed
especially for international projects. This is mainly because principal contractors and sub-
contractors in domestic projects mostly have accumulated experience of working together
on previous (and potantially on future) projects. Since main contractors know features and
performances of many domestic subcontractors, they usually prefer working with the same
subcontractors. However, this is not the case in international projects where the shared expe-
rience is less common and main contractors often have to work with unknown subcontractors
from host countries or totally different third countries. Past relationships between the parties
can be the unique factor in selecting subcontractors in domestic projects, whereas this can
hardly be implemented in international projects especially where main contractor has not
undertaken any project in the host country or city before.
In this context, main contractors need a sound evaluation tool to gain knowledge-based
competition capabilities in the international construction market while choosing subcontrac-
tors. In practice, a few decision makers in main contracting firms choose subcontractors based
on solely cost or a few criteria without employing any evaluation method. The last decision is
often made by top managers based on heuristic techniques combining experience, intuition,
and subjective judgment (Ulubeyli et al. 2010). It means that subcontractor selection deci-
sions are generally a ‘gut feel’ issue. Of course, the same organization can perform differently
even under very similar conditions. However, many contractors have not developed their
well-structured, systematic, and formalized process to prequalify subcontractors, and thus
subcontracting selection process often lacks reliable standards. In this type of tender evalu-
ation, there also appears to be a lack of a realistic working models capable of simultaneously
Technological and Economic Development of Economy, 2016, 22(2): 210–234 213

(i) managing a large number of criteria and the complex selection process, (ii) compiling
inputs for multiple decision-criteria, (iii) reducing the complexity of data handling, (iv) of-
fering computer interaction that makes a model highly flexible to any change in situation
and that facilitates the implementation of the fuzzy sets approach, (v) storing and reporting
all input data, (vi) coping with multi-criteria information, which is imprecise and subjective,
and accommodating non-numerical measures by means of fuzzy reasoning, (vii) giving
an initial idea of how subcontractors perform on each decision criterion and allowing the
main contractor to understand the picture on the strong and weak points of each bidder,
(viii) requiring only one user, who may not be a construction practitioner, to run the model,
(ix) creating a database system for shorter and easier evaluation of future subcontract pack-
ages by a feedback mechanism, and lastly (x) having high practical application. The model
developed in this study gather all these characteristics together.
Moreover, proactive subcontractors who seek added knowledge about a main contractor’s
decision process can use CoSMo as an effective evaluation tool by focusing on the factors that
are perceived to be important to main contractors and by examining their chances of success
to qualify. An understanding of the main contractor’s decision making behavior during the
selection process can equip the subcontractor with the edge required to overcome compet-
itors, or at least to improve their chances of doing so by increasing their chances of being
awarded the sublet work. In brief, CoSMo can motivate the subcontractors continuously to
concentrate on improving characteristics that are of interest to main contractors.
As a result, CoSMo can be used as a main contractor advisory system in choosing sub-
contractors by satisfying main contractor’s needs to reduce the risk involved in the selection
of a subcontractor, to remove subcontractor-based problems during construction, and finally
to complete a construction project successfully (Ulubeyli 2008).

3. Fuzzy multi-criteria decision making

In the present study, the fuzzy sets theory was employed as a solution methodology. This has
several reasons which are explained in a detailed manner below.
During the conventional evaluation process of a decision-making problem, individuals
are often required to give exact or precise numerical assessments with regard to each de-
cision criterion. Although this is achievable, the quantitative representation of subjective
characteristics may impose a heavy burden on the decision maker. Furthermore, scoring
in these numbers may not “truly” reflect the decision maker’s preferences. Humans are not
machines, so they are susceptible to intuitive judgments based on individual knowledge
and experience within domain. Therefore, scoring in linguistic terms is the closest form
of representing human judgment, and decision makers may wish to evaluate criteria by
using linguistic expressions or variables, which facilitates the processing of raw data that
is normally difficult to represent. It is desirable that decision makers not be forced to
provide exact numerical assessments, but rather, should be free to express their judgments
subjectively. The fuzzy sets method introduced by Zadeh (1965) facilitates this, since
asking preferences directly by crisp values for weighting and rating does not sufficiently
accurately represent individual semantic cognition status. Linguistic values are used to
214 S. Ulubeyli, A. Kazaz. Fuzzy multi-criteria decision making model for subcontractor ...

address the vagueness of human thoughts in the selection process. Fuzzy logic is a very
powerful tool that can be used to quantify imprecise data, and has the advantage of math-
ematically represent uncertainty and vagueness. As a conclusion, since knowledge can
be expressed in a more natural way by using fuzzy sets, many engineering and decision
problems can be greatly simplified. One of the most salient features of fuzzy logic is that
subjective information that is available only as a linguistic statement can easily be made
quantitative. The fuzzy inference can also model human judgment by combining quali-
tative and quantitative decision criteria into an aggregate measure.
Therefore, an ideal decision support system for subcontractor prequalification should
have the ability of handling both quantitative and qualitative data, such that rational and
consistent decisions can be made. However, incorporating capability measures of subcon-
tractors into bid evaluation is no easy task. The measurements of qualitative factors can be
described as “an art where subjective judgment, based on an individual’s experience, becomes
an essential part of the process”. The information for these factors is qualitative in nature, and
the assessment and interpretation for such require expert predictive judgment. Typically, it
is necessary to transform the qualitative and quantitative subcontractor’s information into
objective numerical figures via a co-existent treatment. In the modeling process, linguistic
values of selection criteria and evaluation of these criteria for each alternative are converted
to the quantitative format by means of fuzzy sets. This helps decision-makers express linguis-
tic value for criteria. The use of the linguistic values can be justified by the fact that human
decision making involves ambiguity, vagueness, uncertainty, and imprecision. Although it
may be too complex for easy acceptance by industry, this disadvantage was minimized via a
computer-supported system in this study. In addition, fuzzy sets theory as a selection meth-
odology shows an evidence of academic usage (Holt 1998).

4. Fuzzy algorithm of CoSMo

Many simple decision processes are based on a single criterion. Often, however, decisions
should be made in an environment where more than one criterion governs constraints on
the problem, and the value of each of these criteria is different (Lashgari et al. 2012; Yu 2013).
Two primary issues in multi-criteria decision making are to acquire meaningful infor-
mation regarding the satisfaction of the criteria by the various choices or alternatives and to
rank or weight the importance of each of the criteria. The approach illustrated below defines
a decision calculus that requires only “ordinal” information on the ranking of preferences
and importance weights (Yager 1981).
The typical multi-criteria decision problem involves the selection of one alternative, ai,
from a universe of alternatives A given a set, C, of criteria that are important to the decision
maker. Decision maker want to evaluate how well each choice satisfies each criterion, and he/
she wish to combine the weighted criteria into an overall decision function in some plausible
way. This decision function essentially represents a mapping of the alternatives in A to an
ordinal set of ranks. This process naturally requires subjective information from the decision
authority concerning the importance weight of each criterion. Ordinal orderings of these
importance weights are usually the easiest to obtain. Numerical values, ratios, and intervals
Technological and Economic Development of Economy, 2016, 22(2): 210–234 215

expressing the importance weight of each criterion are difficult to extract and, if attempted
and then subsequently altered, can often lead to results inconsistent with the intuition of the
decision maker. To develop this calculus some definitions are required. Define a universe
{ } { }
of n alternatives, A = a1, a2 ,, an and a set of r criteria, C = C1, C2 ,, Cr . Let Ci indicate
the ith criterion. Then the degree of membership of alternative a in Ci , denoted µCi ( a ) is the
degree to which alternative a satisfies the criteria. A decision function that simultaneously
satisfies all of the decision criteria is sought; hence, the decision function, D, is given by the
intersection of all criteria: D = C1 ∩ C2 ∩∩ Cr . Therefore, the grade of membership that
the decision function, D, has for each alternative a is given by:

µ D ( a ) = min µC1 ( a ) , µC2 ( a ) ,, µCr ( a )  . (1)

 
The optimum decision, a*, will then be made by the decision maker according to
the ranking of alternatives in a decreasing order. Let now define a set of preferences,
{P}, which will be constrained to being linear and ordinal. According to Ross (1997),
elements of this preference set can be linguistic values such as none, low, medium, high,
absolute, or perfect; or they could be values on the interval [0, 1]; or they could be values
on any other linearly ordered scale, e.g., [–1, 1], [1, 10], [20, 90] etc. These preferences
will be attached to each of the criteria to quantify the decision maker’s feelings about
the importance that each criterion should have. Let the parameter, bi, be contained on
the set of preferences, {P}, where i = 1, 2,, r . Hence, the decision maker has for each
criterion a measure of how important it is to him/her for a given decision. The decision
function, D, now takes on a more general form when each criterion is associated with a
weight expressing its importance to the decision maker. This function is represented as
the intersection of r-tuples, denoted as a decision measure, M ( Ci , bi ), involving criteria
and preferences,
D = M ( C1 , b1 ) ∩ M ( C2 , b2 ) ∩∩ M ( Cr , br ) . (2)

A key question is what operation should relate each criterion, Ci, and its importance
weight, bi, that preserves the linear ordering required of the preference set, and at the same
time relates the two quantities in a logical way where negation is also accommodated. The
classical implication operator (→) satisfies these requirements. In this implication, the propo-
sition bi is referred to as the hypothesis and the proposition Ci is referred to as the conclusion
(Zadeh 1973). In fact, the relation enables each decision measure, denoted as M(Ci, bi) in
Equation (2), to be calculated mathematically. Hence, the decision measure can be replaced
with the classical implication by means of the implication’s classical equivalent where bi is
the complement of bi,
M ( Ci , bi ) =bi → Ci =bi ∪ Ci . (3)

Justification of the implication as an appropriate measure can be developed using an

intuitive argument. The statement “bi implies Ci” indicates a unique relationship between a
preference and its associated criterion. Whereas various criteria can have the same preference
weighting in a cardinal sense, they will be unique in an ordinal sense even though the equality
216 S. Ulubeyli, A. Kazaz. Fuzzy multi-criteria decision making model for subcontractor ...

situation bi = bj for i ≠ j can exist for some criteria. Ordering will be preserved because
bi ≥ bj will contain the equality case as a subset. Therefore, a reasonable decision model will
be the joint intersection of r decision measures:
r
(
D = ∩ bi ∪ Ci ,
i =1
) (4)

and the optimum solution, a*, will be decided by the decision maker according to the con-
cluding ranking of alternatives. In fact, equation (4) denotes that the model essentially accepts
the “weakest link” approach as suggested by Smithson and Verkuilen (2006). This means that
a chain is only as strong as its weakest link and the minimum (∩) strength of all the links in
the chain governs the strength of the overall chain (Ragin 2000).
This model is intuitive in the following manner. As the ith criterion becomes more
important in the final decision, bi increases, causing bi to decrease, where now Ci ( a ) will
be the value of the decision function, D, representing alternative a. Since the type of input
and output data is group membership, each alternative is ranked in order according to
their membership value. As was used in the utility model for comparing fuzzy numbers
developed by Juang et al. (1987), a subcontractor prequalification ranking index ranging
from 0.1 to 0.9 is obtained; the higher the index, the greater the perceived capability of the
subcontractor. From another perspective, as explained in the above algorithm, membership
values of preference set µbi ( a )  and alternatives’ set for each criterion µCi ( a )  should
   
be determined by decision-maker. It is more suitable that decision-maker expresses his/
her judgment on these values by linguistic variables rather than specific numerical values.
Determining the number of conversion scales which are applied to transform linguistic
terms into fuzzy numbers is generally intuitive (Chen, Hwang 1992). Miller (1965) noted
that the scale of “seven plus or minus two” generates the largest amount of information
from a decision-maker regarding the objectives on the basis of absolute judgments. In
this study, an interval scale of nine was adopted. The choice of nine rating levels of per-
formance is appropriate because more or fewer categories appear to be either too many
or too few for user friendliness and accuracy of evaluation. As shown in Figure 1, nine
linguistic values were used for determining numerical rating values of these variables.
Similarly, nine linguistic weight values are illustrated in Figure 2.

Fig. 1. Linguistic rating terms

Fig. 2. Linguistic weighting terms

Technological and Economic Development of Economy, 2016, 22(2): 210–234 217

A more detailed explanation of the value of this approach should be given. For a particular
criterion, the negation of its preference acts as a barrier such that all ratings of alternatives
below that barrier become equal to the value of that barrier. Here, all distinctions less than
the barrier are disregarded while keeping distinctions above this barrier. However, in the
decision model developed here, this barrier varies, depending upon the preference of the
criterion to the decision maker. The more important is the criterion, the lower is the barrier,
and thus the more levels of distinction there are. As a criterion becomes less important, the
distinction barrier increases, which lessens the penalty to the criterion. In the limit, if the
criterion becomes totally unimportant, then the barrier is raised to its highest level and all
alternatives are given the same weight and no distinction is made. Conversely, if the criterion
becomes the most important, all distinctions remain. In sum, the more important a criterion
is in the decision process, the more significant its effect on the decision function, D (Yager
1981; Ross 1997).

5. The structure of CoSMo

Decision support systems can be defined as “computer-aided systems designed to assist

decision-makers to make better, faster, and/or cheaper decisions”. They organize evaluation
processes and help to make the processes less sophisticated, less time- and effort-con-
suming, more efficient in analyzing tender-data, and easier to execute. These systems
facilitate computations through computer-supported models and ensure that the models
are put into practice especially where the complicated algorithms and calculations such
as fuzzy sets are in use.
In order to efficiently program and run CoSMo, the computational system was created
using Visual Basic for Application (VBA) and MS Access, which are Windows-standard
computing softwares. The platform of Windows XP was adopted for developing CoSMo as
the programming application has the advantage of good capacity for model programming
and database management. It is also an important advantage that CoSMo occupies only a
space of 30–32 megabytes in the hard disc, together with recorded data. The release package
for the system was designed so that it can be easily installed to a Windows system and
can run independently without the need for additional configurations. CoSMo can also
be used in computers where Microsoft Office has not been set up. The most important
advantage of employing VBA and MS Access is their familiarity for most users. VBA was
used to code the detailed processes of CoSMo, to automatically handle tasks, to render
a high interface, and to offer friendly interaction with users. “If … Then … Else” type
codes were produced in the VBA environment. In other words, the database’s interface
was designed with VBA to make easier the interaction between CoSMo and users, hid-
ing the calculations and operations from the user’s eyes. MS Access was selected in the
present research effort, because it is a good storing and reporting tool. It also applies to
VBA codes as modules, and user forms and reports supported by these modules work
as menus that execute the proposal evaluation process. The main menu that shows and
manages the modules of this program is illustrated in Figure 3.
218 S. Ulubeyli, A. Kazaz. Fuzzy multi-criteria decision making model for subcontractor ...

Fig. 3. Main menu of CoSMo

The entire algorithm presented in Section 5 is repeated in the same way in all phases of
CoSMo, i.e., “short-listing”, “negotiation”, “final selection”, and “outcome”, to reach final values.
In using CoSMo, a main contractor needs to collect information from bidding subcontractors
in referring to the parameters. After inputting the information into CoSMo, the system allows
each subcontractor’s final value to be calculated. In other words, after allocating rating and
weight values for criteria of “short-listing”, “negotiation”, and “final selection”, three interme-
diate values are obtained for each alternative. In the “outcome” step, these three values are
automatically included as ratings of previous three stages. Then, weight values of them are
assessed by decision maker(s) and final values are calculated.

6. Procedures in the subcontractor selection process

As demonstrated in Figure 4, the forms “short-listing”, “negotiation”, “final selection”, and

“outcome”, which are perceived as separate objects by the program, operate as data input
modules, while the “briefing” report runs as the resultant output module. Consequently, it
was intended in this study to develop specific modules with a special focus on subcontractor
selection for international projects, where the evaluation parameters are more in number and
complexity due to availability of mostly unknown bidders, and therefore decision support
is even more valuable.
In general, all prequalification systems have the same basic steps: develop the criteria,
gather the related data, apply the data to criteria, and decide whether to prequalify or remove
the candidate. This procedure was also met by CoSMo using a three-step decision hierarchy.
This three-stage process including “short-listing”, “negotiation”, and “final selection”, ensures
that subcontract packages do not reflect unwarranted variations in likely performance of
subcontractors. Toward this aim, subcontractors must qualify for each stage. The decision
maker introduces a cut-off point for prequalification as follows: the subcontractors whose
membership values equals lower than “moderate” for “the sum of short-listing” is not con-
sidered for the negotiation and final selection stages. Namely, short-listing scores below this
Technological and Economic Development of Economy, 2016, 22(2): 210–234 219

Entrance Menu Exit

Module 1

Short-listing

Module 5 Module 2

Briefing Negotiation

CoSMo
Database

Module 4 Module 3

Outcome Final
Selection

: Input module : Output module

Fig. 4. Operating mechanism of CoSMo

pre-determined threshold are deemed non-responsive. This frontier acts as a fixed limit from
which all subcontractors can be compared, and serves as a measure of the quality of subcon-
tractors vying for the project. The baseline is also believed to reflect the industry’s standard
in general. In practice, however, it is possible that no one passes the threshold. In such a case,
main contractor may review his selection strategy and reduce the current standard.
Figure 5 shows the application hierarchy of the first stage, i.e., “short-listing”. First, an
in-house committee in the name of the main contracting company invites subcontractors for
preliminary negotiation and requests their formal documents concerning past experience,
past performance, financial strength, workload, safety record, litigation history, and location
of home office to evaluate them and assign corresponding ratings. In the document about
past performance, randomly-selected some information may be checked by communicating
with the related past main contractors. Formal relationship, personal relationship, and repu-
tation are the other criteria to be considered in this stage. Ratings of formal relationship and
personal relationship are determined according to the preliminary negotiation while those
of reputation are assigned after controlling reputable past projects undertaken by applicants.
After compiling all of the information about these ten criteria, the committee makes a meeting
to determine rating scores and project-specific weights of them. It is suggested that decisions
on these values are taken by consensus rather than by majority. The required explanations on
these criteria should be added to CoSMo to store for using in future evaluations. The fuzzy
algorithm is then applied automatically and, as mentioned above, those who have scores
greater than a pre-determined threshold qualify. These bidders are sent detailed tender doc-
uments and project drawings to give their cost- and time-based proposals.
Figure 6 illustrates the selection process followed in the second and third stages,
i.e., “negotiation” and “final selection”. Here, a detailed negotiation is held between the
committee and subcontractors who have passed the first stage. Under the “negotiation”
stage, the committee tests subcontractors on seven criteria such as knowledge of project,
reliability, selfless attitudes, ability to solve problems, enthusiasm for the project, quality
220 S. Ulubeyli, A. Kazaz. Fuzzy multi-criteria decision making model for subcontractor ...

Stage 1

Invite subcontractors to
Related main
bidding process
contractors and
other resources

Obtain information about

short-listing criteria from
subcontractors
Update Update

Committee meeting

Determine weights of criteria

CoSMo Crit. 1
database .
.
Crit. 10

Decide rating values of

criteria

Note required explanations if

Do not take to Stage 2

any

Fuzzy algorithm

No Qualify?

Yes

Approved subcontractors
Subcon. 1
Subcon. 2
.
.
Subcon. n
Send project and tender
documents to subcontractors

Stages 2 and 3

Fig. 5. Flow chart of the “short-listing” procedure

Technological and Economic Development of Economy, 2016, 22(2): 210–234 221

Stages 2 and 3

Invite subcontractors to
negotiate

Take proposals from subcontractors

Information flow / update

Information flow / update
Information flow / update

Committee-subcontractor
meetings

Discussion between
committee members after the
meetings
Negotiation Final selection

CoSMo
database Determine weights of criteria
Crit. 1 Crit. 1
. .
. .
Crit. 7 Crit. 7

Decide rating values of

criteria

Note required explanations if

any

Fuzzy algorithm
Fix weights of three stages

Fuzzy algorithm Subcon. 1

.
.
Subcon. n

Rank final results Subcontractor X

Subcontractor selection
by the committee

Fig. 6. Flow diagram of the “negotiation” and “final selection” procedures

222 S. Ulubeyli, A. Kazaz. Fuzzy multi-criteria decision making model for subcontractor ...

awareness, and level of communication. In fact, this stage is required to make evaluations
on criteria of “negotiation” and “final selection” because (i) it is both time- and effort-con-
suming for the committee to take into account all of the candidate firms without applying
any pre-qualification stage (“short-listing” in this model) and to negotiate with them in a
detailed manner in the “short-listing” stage and (ii) “negotiation” and “final selection” stages
focus on the detailed information about project-specific characteristics while “short-listing”
is directly related with the core competencies of candidates. In terms of the “final selection”
stage, there are seven more evaluation criteria such as price, technical personnel, labor,
equipment, payment plan, amount of subcontracting, and amount of compensation for
delay. After the meeting, committee members discuss for the ratings and weights of the cri-
teria of “negotiation” and “final selection”. In this process, necessary explanations should be
noted to CoSMo especially regarding “final selection” criteria which are tangible in nature.
Then, the fuzzy algorithm is automatically applied for “negotiation” and “final selection”. For
subcontractors who have qualified after “short-listing”, the sum values of three stages are
automatically appeared in the “outcome” form as rating scores of three stages. Finally, the
committee determines weights of these three stages for the last automatic fuzzy calculation.
This procedure ensures that high scores in one, perhaps less important area, do not hide a
low score in another category. Finally, in the “briefing” form, the concluding results obtained
in the “outcome” form are ranked and reported in a decreasing order. In this context, CoSMo
gives an opportunity of making the last decision to the committee instead of calculating the
top subcontractor only, and the results calculated for all candidates are reported in a ranking
order. Accordingly, there is no obligation to select the top bidder. The committee can take a
responsibility to award the second top bidder after investigating why it had a lower “outcome”
value. However, if the difference between final results is larger than one scale, it is not rec-
ommended that the committee takes the initiative to choose the second top subcontractor.

7. Case study

To gain a better understanding of the current practice of subcontractor selection, a case of

international construction works was considered. Thus, the applicability of CoSMo was also
validated by conducting a real world construction project. The main contractor in this case
study was from Turkey and has had experience in the international construction market for
decades. The project for which the firm would choose a subcontractor was a multi-objective
residence complex in Moscow, Russia. It contained a multi-storey building complex which
was composed of luxurious office and dwelling areas, a car park, and a technical annexe.
As usual, the main contractor would divide the project into several packages, and give each
part to a subcontractor. In this case study, the subcontractor selection process for basic con-
struction works was examined. A total of six subcontractors were interested in this package.
From here, companies are denoted by SC1, SC2, SC3, SC4, SC5, and SC6 for confidentiality.
SC1, SC2, and SC3 were the applicants for the package while SC4, SC5, and SC6 were the
invited companies. SC1, SC4, and SC6 were Turkish firms, SC2 was a Russian firm, SC3 was
a Ukrainian firm, and SC5 was an American firm.
Technological and Economic Development of Economy, 2016, 22(2): 210–234 223

Linguistic ratings of 24 criteria and three stages were presented in Table 1 as a whole. In
addition to these scores, their linguistic weights were also given in brackets. According to
the data, the main contractor perceived “past experience”, “selfless attitudes”, and “enthusi-
asm for the project” as the most predominant criteria in his own selection strategy, while
“amount of compensation for delay” was identified as the least important parameter. It was
interesting that in the “final selection” stage there was no criterion perceived as “extremely
high” in importance whereas in the “negotiation” stage there were two such criteria. This
obviously showed how much important the intangible criteria could be. However, when all
stages were taken into account together, “negotiation” was evaluated as the least important
stage while “short-listing” was fixed as the most significant one. Nevertheless, importance
levels placed on three stages were all above “high”. The main contracting company established
a subcontractor prequalification committee including a member of board of directors, project
coordinator, proposal/tender manager, project manager, and site manager.
From previous experience with various subcontractors, the committee first reviewed the
short-listing information about candidates and decided on the prequalified subcontractors.
As an example, data and the result of “short-listing” for SC1 were illustrated in Figure 7.
For the “short-listing” stage, the selection problem including alternatives (A) and criteria
(C) was set up as follows:

A = {SC1, SC 2, SC 3, SC 4, SC 5, SC 6} , (5)

C = {S1, S2, S3, S 4, S5, S6, S7, S8, S9, S10} . (6)

In this stage, the committee rated the subcontractors with respect to the ten criteria,
as shown in Table 1. These ratings are fuzzy sets expressed in Zadeh’s (1965) notation. For
instance, the fuzzy set of S1 (“past experience”) can be given as follows:

M P P FG VG FG
S1 = + + + + + . (7)
SC1 SC 2 SC 3 SC 4 SC 5 SC 6

Fig. 7. Screen of the “short-listing” module

224 S. Ulubeyli, A. Kazaz. Fuzzy multi-criteria decision making model for subcontractor ...

Table 1. Linguistic ratings/weights of criteria and results of selection stages

Com­ S1 S2 S3 S4 S5 S6 S7 S8 S9 S10
Stage
pany (EH) (VH) (H) (FH) (H) (H) (A) (L) (H) (FL)
SC1 M M M M G G G P VG M
SC2 P G M M FG P P E M E
Short- SC3 P FG M M VG VP VP E M FG
list­ing SC4 FG G G G G FG G E E M
SC5 VG FG FG E FG E E E VG VG
SC6 FG FG VG VG FG VG VG E VG M
N1 N2 N3 N4 N5 N6 N7
Com­pany
(H) (FH) (EH) (VH) (EH) (VH) (A)
Nego­ SC1 VG FG VG M M M G
tia­tion SC4 E FG VG FG G FG VG
SC5 E FG FG FG VG VG FG
SC6 E VG VG VG E VG VG
F1 F2 F3 F4 F5 F6 F7
Com­pany
(FH) (VH) (FH) (H) (A) (VL) (EL)
Final SC1 FG M G M M FG M
selec­
SC4 G FG G G E FP M
tion
SC5 VG VG FG VG FG N VG
SC6 FG FG VG FG FG N G
SO
SS SN SF
Com­­pany (Concluding
(EH) (FH) (VH)
result)
SC1 M M M M
Out­ SC2 P Out of evalua­tion Out of evalua­tion Unqualified
come
SC3 P Out of evalua­tion Out of evalua­tion Unqualified
SC4 G G G G
SC5 FG FG FG FG
SC6 FG VG FG FG
Note: S1: Past experience; S2: Past performance; S3: Formal relationship; S4: Financial strength; S5: Work-
load; S6: Safety records; S7: Reputation; S8: Litigation history; S9: Personal relationship; S10: Location of
home office.
N1: Knowledge of project; N2: Reliability; N3: Selfless attitudes; N4: Ability to solve problems; N5: Enthusi-
asm for the project; N6: Quality awareness; N7: Level of communication.
F1: Price; F2: Technical personnel; F3: Labor; F4: Equipment; F5: Payment plan; F6: Amount of subcontract-
ing; F7: Amount of compensation for delay.
SS: The sum of short-listing; SN: The sum of negotiation; SF: The sum of final selection; SO: The sum of
outcome.
N: None; VP: Very poor; FP: Fairly poor; P: Poor; M: Moderate; G: Good; FG: Fairly good; VG: Very good;
E: Excellent.
EL: Extremely low; VL: Very low; FL: Fairly low; L: Low; A: Average; H: High; FH: Fairly high; VH: Very
high; EH: Extremely high.
Technological and Economic Development of Economy, 2016, 22(2): 210–234 225

This kind of membership functions for each of the alternatives was formed in further
calculations. The committee listed its preferences (weights), the set of “b”, for each of the ten
criteria. These were also shown in Table 1 in brackets under criteria symbols. From these
preference values, the following calculations resulted:

=b {EH
= ,VH , H , FH , H , H , A, L, H , FL}
{0.9, 0.8, 0.6, 0.7, 0.6, 0.6, 0.5, 0.4, 0.6, 0.3} ; (8)

b = {0.1, 0.2, 0.4, 0.3, 0.4, 0.4, 0.5, 0.6, 0.4, 0.7} . (9)

Solving the problem for SC1 as an example, the following result was obtained:

( )∩(b ∪S2)∩(b ∪S3)∩(b ∪S4 )∩(b ∪S5)∩

D ( SC1) = b1 ∪S1 2 3 4 5

(b ∪S6)∩(b ∪S7 )∩(b ∪S8)∩(b ∪S9)∩(b ∪S10) =

6 7 8 9 10

( 0.1∪0.5) ∩ ( 0.2∪0.5) ∩( 0.4∪0.5) ∩( 0.3∪0.5) ∩( 0.4∪0.6 ) ∩

( 0.4∪0.6 ) ∩( 0.5∪0.6 ) ∩( 0.6∪0.4 ) ∩( 0.4∪0.8 ) ∩( 0.7∪0.5) =
0.5 ∩ 0.5 ∩ 0.5 ∩ 0.5 ∩ 0.6 ∩ 0.6 ∩ 0.6 ∩ 0.6 ∩ 0.8 ∩ 0.7 = 0.5 ( M ) . (10)

At the end of this “short-listing” stage, the resultant values of SC2 and SC3 were calculated
as “poor”, i.e., lower than the threshold “moderate”. Therefore, these subcontractors were
neither evaluated nor screened in the remaining two stages and were reported as “unqual-
ified” in the “briefing” form. In this context, SC1, SC4, SC5, and SC6 were the bidders that
could pass to “negotiation” and “final selection”. To be graphical examples, data and results
of “negotiation” and “final selection” for SC1 were respectively illustrated in Figures 8 and 9.
For the “negotiation” stage, the problem including alternatives (A) and criteria (C) was
formed as follows:
A = {SC1, SC 4, SC 5, SC 6}; (11)

C = {N 1, N 2, N 3, N 4, N 5, N 6, N 7}. (12)

Here, the committee rated these four subcontractors with respect to the seven criteria
(Table 1). The fuzzy set of N1 (“knowledge of project”) can be given as an example:

VG E E E
N1 = + + + . (13)
SC1 SC 4 SC 5 SC 6

The committee listed its preferences (the set of “b”) for each of the seven criteria (Table 1).
From these preference values, the following calculations resulted:

b = {H , FH , EH ,VH , EH ,VH , A} = 0.6, 0.7, 0.9, 0.8, 0.9, 0.8, 0.5 ; (14)
b = {0.4, 0.3, 0.1, 0.2, 0.1, 0.2, 0.5}. (15)
226 S. Ulubeyli, A. Kazaz. Fuzzy multi-criteria decision making model for subcontractor ...

Solving the problem for SC1 as an instance, the following result was calculated:

( )∩(b ∪N 2)∩(b ∪N 3)∩(b ∪N 4 )∩

D ( SC1) = b1 ∪N 1 2 3 4

(b ∪N 5)∩(b ∪N 6)∩(b ∪N 7 ) =
5 6 7

( 0.4 ∪ 0.8 ) ∩ ( 0.3 ∪ 0.7 ) ∩ ( 0.1∪ 0.8 ) ∩ ( 0.2 ∪ 0.5 ) ∩

( 0.1∪ 0.5 ) ∩ ( 0.2 ∪ 0.5 ) ∩ ( 0.5 ∪ 0.6 ) =
0.8 ∩ 0.7 ∩ 0.8 ∩ 0.5 ∩ 0.5 ∩ 0.5 ∩ 0.6 = 0.5 ( M ). (16)

Similarly, for the “final selection” stage, the problem including alternatives (A) and criteria
(C) was formed as follows:
A = {SC1, SC 4, SC 5, SC 6}; (17)

C = {F1, F 2, F 3, F 4, F 5, F 6, F 7}. (18)

Fig. 8. Screen of the “negotiation” module

Fig. 9. Screen of the “final selection” module

Technological and Economic Development of Economy, 2016, 22(2): 210–234 227

The committee rated these four subcontractors with respect to the seven criteria (Table 1).
The fuzzy set of F1 (“price”) can be given as follows:
FG G VG FG
F1 = + + + . (19)
SC1 SC 4 SC 5 SC 6
The committee listed its preferences, the set of “b”, for each of the seven criteria (Table 1).
From these preference values, the following calculations resulted:

b = {FH ,VH , FH , H , A,VL, EL} = {0.7, 0.8, 0.7, 0.6, 0.5, 0.2, 0.1} ; (20)

b = {0.3, 0.2, 0.3, 0.4, 0.5, 0.8, 0.9} . (21)

Solving the problem for SC1 as an example, the following result was obtained:

(
D ( SC1) = b1 ∪F1 )∩(b ∪F 2)∩(b ∪F 3)∩(b ∪F 4 )∩
2 3 4

(b ∪F 5)∩(b ∪F 6)∩(b ∪F 7 ) =
5 6 7

( 0.3∪0.7 ) ∩ ( 0.2∪0.5) ∩( 0.3∪0.6 ) ∩( 0.4∪0.5) ∩

( 0.5∪0.5) ∩( 0.8∪0.7 ) ∩( 0.9∪0.5) =
0.7 ∩ 0.5 ∩ 0.6 ∩ 0.5 ∩ 0.5 ∩ 0.8 ∩ 0.9 = 0.5 ( M ) . (22)

For each subcontractor, the results of the “outcome” stage were calculated. As an instance,
data and the result of “outcome” for SC1 were illustrated in Figure 10.
For the “outcome” stage, the selection problem including alternatives (A) and criteria (C)
was set up as follows:
A = {SC1, SC 4, SC 5, SC 6} ; (23)

C = {SS, SN , SF} . (24)

Fig. 10. Screen of the “outcome” module

 228 S. Ulubeyli, A. Kazaz. Fuzzy multi-criteria decision making model for subcontractor ...

As criteria in this step are selection stages, their rating values were automatically taken
from stage results for each subcontractor (Table 1). For example, the fuzzy set of SS (“the
sum of short-listing”) can be given as follows:
M G FG FG
SS = + + + . (25)
SC1 SC 4 SC 5 SC 6
The committee listed its preferences, the set of “b”, for each of the three stages (Table 1).
From these preference values, the following calculations resulted:
=b {EH
= , FH ,VH } {0.9, 0.7, 0.8} ; (26)

b = {0.1, 0.3, 0.2}. (27)

Solving the problem for SC1 as an instance, the following result was calculated:

=D ( SC1) (b=
∪SS ) ∩ (b ∪SN ) ∩(b ∪SF )
1 2 3

( 0.1∪0.5) ∩ ( 0.3∪0.5) ∩( 0.2∪0.5) =

0.5 ∩ 0.5 ∩ 0.5 = 0.5 ( M ) . (28)

In the “briefing” module, the concluding rankings shown in Figure 11 were obtained. In
conclusion, the committee determined SC5 and SC6 as the top two bidders.

Fig. 11. Screen of the “briefing” module

Technological and Economic Development of Economy, 2016, 22(2): 210–234 229

Fig. 12. Screen of the “firm details” page in the “information on firms” module

When desired, as shown in Figure 12, each subcontractor’s partial scores can be viewed
as a whole in one page (the “firm details” page) from the “information on firms” module in
the main menu presented in Figure 3.
As there were two equal winners of the subcontracting package, an important issue
in practice was taken into account by the committee. Since subcontractors in the same
or successive activities work in an interactive manner with each other, cooperation and
communication between them should be problem-free. This issue may likely turn to a vital
problem in international construction projects where cultural differences between crews
can have utmost importance. The traditional way to subcontract a project does not allow
for thorough evaluation of the interdependence of various participating entities in each
project, and generally cuts each specific project into several completely independent sub-
contracts. However, a general contractor should consider the entire subcontracting supply
chain of projects because what is individually optimal combination may not necessarily
result in the best global outcome. Therefore, in order to reach better harmony between
real physical constructers of projects and thereby to obtain a better-quality final product,
subcontractors who have worked together in previous projects can be employed. While
using too many subcontractors and second-tier subcontractors in a large-scale construction
project, supervising and controlling become more difficult for principal contractors and
extra conflicts can rise between crews. Olsson (1998) explains that several main contrac-
tors in Sweden use a special subcontract coordinator before and during construction in
an effort to reduce problems.
230 S. Ulubeyli, A. Kazaz. Fuzzy multi-criteria decision making model for subcontractor ...

As a result, whether healthy working environment between the selected two firms and
the other potential subcontractors that might undertake the remaining work packages of the
project could be established was taken into account by the committee. After investigating
past experiences of subcontractors who have applied to be awarded for several packages of
final construction works, the committee determined that SC6 has worked with the majority
of these applicants in harmony in past international projects. Consequently, the committee
decided to employ SC6 instead of SC5, and presented the decision letter and the attached
printout reports of each stage of CoSMo to the approval of Board of Directors. At last, SC6
was awarded after the approval.

Conclusions

In this study, a computer-aided subcontractor selection model called CoSMo that can help a
project contractor in conducting a prequalification assessment among various combinations of
on-list subcontractors was introduced. The mission of this computational model is to emphasize
the importance and necessity of the subcontractor selection problem in international construc-
tion projects. The model presented in this study can be described as a dynamic, reliable, and
practical automation system that was created to choose the most suitable subcontractor for a
specific project package rather than the “best” subcontracting firm in absolute terms. Selecting
the most appropriate tender with high confidence, in turn, help in reducing the effort and time
consumed in the evaluation process. General contractors can thus concentrate on attracting
professional subcontractors who will bring and produce quality work, which will ultimately
have a significant impact on improving the main contractors’ business reputation. It should
also be noted that the results obtained by CoSMo are completely human-dependent and that
each decision strategy is subject to biases of the decision maker(s). In this respect, the results
presented provide insight into how decision makers view their decision factors in the subcon-
tractor prequalification. Subcontractor selection inputs themselves are critical procurement
aspects that should be tailored to match project objectives despite the fact that decision-makers
may give different weights of importance to the same criterion and give different values of a
criterion for each subcontractor over time. Toward this aim, in recognition of the changes of
preference in relation to a particular situation, CoSMo allows decision-makers to change weights
and ratings of criteria and stages as required. This option provides flexibility for CoSMo. In
addition, since different decision models may provide different solutions, CoSMo can be used as
a benchmark dictionary by potential researchers in the future. However, in this study, it cannot
be claimed that perfect or the best decision-making model was found out, because all of deci-
sion methodologies are based on particular theoretical acceptances together with limitations
and each aid has some disadvantages in arriving at a solution. Nevertheless, users should be
aware of limitations of any technique. For the present study, it is recommended that CoSMo
users may exercise different scenarios based on their changeable or indecisive judgments on
ratings and weights. The fact that the “briefing” module in CoSMo shows the final ranking of
candidates instead of giving only one result addresses this notion. The novel architecture for
the subcontractor selection in a fuzzy environment presented can be easily extended to other
managerial and industrial decision-making problems.
Technological and Economic Development of Economy, 2016, 22(2): 210–234 231

As an on-going research, a detailed computer-supported subcontractor resource module

where CoSMo will be integrated into a project management software as a component is still
being developed. Therefore, this study solely proposes a preliminary model to establish a
standardized environment to accelerate the subcontracting procurement process of a con-
struction project. In this module to be developed, when the decision maker picks the awarded
subcontractor on the CoSMo screen, all the resource information (labor and equipment) of
this firm will be automatically transferred to the module. In the main project program, time
and cost schedules of project activities that will be executed by the subcontractor will then
be formed instantly. Hence, both the subcontractor selection and the related planning efforts
will be carried out in one time.

Acknowledgements

The authors gratefully acknowledge the managers of the surveyed company for their gen-
erous collaboration and contributions. The authors also thank the anonymous referees for
their constructive comments which have helped in improving the paper. Finally, the authors
would like to thank the financial supports provided by the Committees on Research Grants
of Bulent Ecevit University and Akdeniz University.

References
Albino, V.; Garavelli, A. C. 1998. A neural network application to subcontractor rating in construction
firms, International Journal of Project Management 16(1): 9–14.
http://dx.doi.org/10.1016/S0263-7863(97)00007-0
Arditi, D.; Chotibhongs, R. 2005. Issues in subcontracting practice, Journal of Construction Engineering
and Management 131(8): 866–876. http://dx.doi.org/10.1061/(ASCE)0733-9364(2005)131:8(866)
Arslan, G. 2012. Web-based contractor evaluation system for mass-housing projects in Turkey, Journal of
Civil Engineering and Management 18(3): 323–334. http://dx.doi.org/10.3846/13923730.2012.698892
Arslan, G.; Kivrak, S.; Birgonul, M. T.; Dikmen, I. 2008. Improving sub-contractor selection process
in construction projects: web-based sub-contractor evaluation system (WEBSES), Automation in
Construction 17(4): 480–488. http://dx.doi.org/10.1016/j.autcon.2007.08.004
Brauers, W. K. M.; Zavadskas, E. K.; Turskis, Z.; Vilutiene, T. 2008. Multi-objective contractor’s ranking
by applying the MOORA method, Journal of Business Economics and Management 9(4): 245–255.
http://dx.doi.org/10.3846/1611-1699.2008.9.245-255
Chen, S. J.; Hwang, C. L. 1992. Fuzzy multiple attribute decision-making method and applications. New
York: Springer-Verlag. http://dx.doi.org/10.1007/978-3-642-46768-4
Cheng, M. Y.; Kang, S. T. 2012. Integrated fuzzy preference relations with decision utilities for construction
contractor selection, Journal of the Chinese Institute of Engineers 35(8): 1051–1063.
http://dx.doi.org/10.1080/02533839.2012.708510
Chiang, Y. H. 2009. Subcontracting and its ramifications: a survey of the building industry in Hong Kong,
International Journal of Project Management 27(1): 80–88.
http://dx.doi.org/10.1016/j.ijproman.2008.01.005
Choudhry, R. M.; Hinze, J. W.; Arshad, M.; Gabriel, H. F. 2012. Subcontracting practices in the construc-
tion industry of Pakistan, Journal of Construction Engineering and Management 138(12): 1353–1359.
http://dx.doi.org/10.1061/(ASCE)CO.1943-7862.0000562
232 S. Ulubeyli, A. Kazaz. Fuzzy multi-criteria decision making model for subcontractor ...

De Silva, D. G.; Kosmopoulou, G.; Lamarche, C. 2012. Survival of contractors with previous subcontracting
experience, Economic Letters 117: 7–9. http://dx.doi.org/10.1016/j.econlet.2012.04.071
Elton, D. J.; Juang, C. H.; Russell, J. S. 1994. Contractor prequalification using fuzzy sets, Civil Engineering
Systems 11: 1–17. http://dx.doi.org/10.1080/02630259408970134
Grasso, B.; Rasdorf, W.; Bridgers, M. 2008. Nature and extent of domestic construction program out-
sourcing, Journal of Construction Engineering and Management 134(12): 1002–1010.
http://dx.doi.org/10.1061/(ASCE)0733-9364(2008)134:12(1002)
Hinze, J.; Tracey, A. 1994. The contractor-subcontractor relationship: the subcontractor’s point of view,
Journal of Construction Engineering and Management 120(2): 274–287.
http://dx.doi.org/10.1061/(ASCE)0733-9364(1994)120:2(274)
Holt, G. D. 1998. Which contractor selection methodology?, International Journal of Project Management
16(3): 153–164. http://dx.doi.org/10.1016/S0263-7863(97)00035-5
Ip, W. H.; Yung, K. L.; Wang, D. 2004. A branch and bound algorithm for sub-contractor selection
in agile manufacturing environment, International Journal of Production Economics 87: 195–205.
http://dx.doi.org/10.1016/S0925-5273(03)00125-7
Juang, C. H.; Burati, J. L.; Kalindindi, S. N. 1987. A fuzzy system for bid proposal evaluation using micro-
computers, Civil Engineering Systems 4(3): 124–130. http://dx.doi.org/10.1080/02630258708970474
Kazaz, A.; Ulubeyli, S.; Tuncbilekli, N. A. 2012. Causes of delays in construction projects in Turkey, Journal
of Civil Engineering and Management 18(3): 426–435. http://dx.doi.org/10.3846/13923730.2012.698913
Kumaraswamy, M. M.; Matthews, J. D. 2000. Improved subcontractor selection employing partnering
principles, Journal of Management in Engineering 16(3): 47–57.
http://dx.doi.org/10.1061/(ASCE)0742-597X(2000)16:3(47)
Lashgari, A.; Yazdani-Chamzini, A.; Fouladgar, M. M.; Zavadskas, E. K.; Shafiee, S.; Abbate, N. 2012.
Equipment selection using fuzzy multi criteria decision making model: key study of Gole Gohar Iron
Mine, Inzinerine Ekonomika – Engineering Economics 23(2): 125–136.
Lingard, H. C.; Cooke, T.; Blismas, N. 2010. Safety climate in conditions of construction subcontracting:
a multi-level analysis, Construction Management and Economics 28: 813–825.
http://dx.doi.org/10.1080/01446190903480035
Mbachu, J. 2008. Conceptual framework for the assessment of subcontractors’ eligibility and performance
in the construction industry, Construction Management and Economics 26: 471–484.
http://dx.doi.org/10.1080/01446190801918730
Miller, G. A. 1965. The magic number seven, plus or minus seven, Psychological Review 63: 81–97.
http://dx.doi.org/10.1037/h0043158
Mitkus, S.; Trinkuniene, E. 2008. Reasoned decisions in construction contracts evaluation, Technological
and Economic Development of Economy 14(3): 402–416.
http://dx.doi.org/10.3846/1392-8619.2008.14.402-416
Ng, S. T.; Luu, C. D. T. 2008. Modeling subcontractor registration decisions through case-based reasoning
approach, Automation in Construction 17(7): 873–881. http://dx.doi.org/10.1016/j.autcon.2008.02.015
Ng, S. T.; Tang, Z. 2010. Labour-intensive construction sub-contractors: their critical success factors, Inter-
national Journal of Project Management 28: 732–740. http://dx.doi.org/10.1016/j.ijproman.2009.11.005
Nguyen, V. U. 1985. Tender evaluation by fuzzy sets, Journal of Construction Engineering and Management
111(3): 231–243. http://dx.doi.org/10.1061/(ASCE)0733-9364(1985)111:3(231)
Okoroh, M. I.; Torrance, V. B. 1999. A model for subcontractor selection in refurbishment projects, Con-
struction Management and Economics 17(3): 315–327. http://dx.doi.org/10.1080/014461999371529
Olsson, R. 1998. Subcontract coordination in construction, International Journal of Production Economics
56–57: 503–509. http://dx.doi.org/10.1016/S0925-5273(97)00024-8
Technological and Economic Development of Economy, 2016, 22(2): 210–234 233

Paek, J. H.; Lee, Y. W.; Napier, T. R. 1992. Selection of design/build proposal using fuzzy-logic system,
Journal of Construction Engineering and Management 118(2): 303–317.
http://dx.doi.org/10.1061/(ASCE)0733-9364(1992)118:2(303)
Plebankiewicz, E. 2009. Contractor prequalification model using fuzzy sets, Journal of Civil Engineering
and Management 15(4): 377–385. http://dx.doi.org/10.3846/1392-3730.2009.15.377-385
Plebankiewicz, E. 2010. Construction contractor prequalification from Polish clients’ perspective, Journal
of Civil Engineering and Management 16(1): 57–64. http://dx.doi.org/10.3846/jcem.2010.05
Podvezko, V.; Mitkus, S.; Trinkuniene, E. 2010. Complex evaluation of contracts for construction, Journal
of Civil Engineering and Management 16(2): 287–297. http://dx.doi.org/10.3846/jcem.2010.33
Poyhonen, M. A.; Hamalainen, R. P.; Salo, A. A. 1997. An experiment on the numerical modelling of
verbal ratio statements, Journal of Multi-Criteria Decision Analysis 6(1): 1–10.
http://dx.doi.org/10.1002/(SICI)1099-1360(199701)6:1<1::AID-MCDA111>3.0.CO;2-W
Ragin, C. C. 2000. Fuzzy-set social science. Chicago: University of Chicago Press.
Ross, T. J. 1997. Fuzzy logic with engineering applications. Singapore: McGraw Hill.
San Cristobal, J. R. 2012. Contractor selection using multicriteria decision-making methods, Journal of
Construction Engineering and Management 138(6): 751–758.
http://dx.doi.org/10.1061/(ASCE)CO.1943-7862.0000488
Shash, A. A. 1998. Bidding practices of subcontractors in Colorado, Journal of Construction Engineering
and Management 124(3): 219–225. http://dx.doi.org/10.1061/(ASCE)0733-9364(1998)124:3(219)
Singh, D.; Tiong, R. L. K. 2005. A fuzzy decision framework for contractor selection, Journal of Construc-
tion Engineering and Management 131(1): 62–70.
http://dx.doi.org/10.1061/(ASCE)0733-9364(2005)131:1(62)
Smithson, M.; Verkuilen, J. 2006. Fuzzy set theory: applications in the social sciences. California: Sage.
Tam, V. W. Y.; Shen, L. Y.; Kong, J. S. Y. 2011. Impacts of multi-layer chain subcontracting on project
management performance, International Journal of Project Management 29: 108–116.
http://dx.doi.org/10.1016/j.ijproman.2010.01.005
Tserng, H. P.; Lin, P. H. 2002. An accelerated subcontracting and procuring model for construction
projects, Automation in Construction 11: 105–125. http://dx.doi.org/10.1016/S0926-5805(01)00056-5
Turskis, Z. 2008. Multi-attribute contractors ranking method by applying ordering of feasible alternatives
of solutions in terms of preferability technique, Technological and Economic Development of Economy
14(2): 224–239. http://dx.doi.org/10.3846/1392-8619.2008.14.224-239
Ulubeyli, S. 2008. Fuzzy multiple criteria decision making model for subcontractor selection in international
construction projects: PhD Thesis. Istanbul University, Turkey.
Ulubeyli, S.; Manisali, E.; Kazaz, A. 2010. Subcontractor selection practices in international construction
projects, Journal of Civil Engineering and Management 16(1): 47–56.
http://dx.doi.org/10.3846/jcem.2010.04
Vahdani, B.; Mousavi, S. M.; Hashemi, H.; Mousakhani, M.; Moghaddam, R. T. 2013. A new compromise
solution method for fuzzy group decision-making problems with an application to the contractor
selection, Engineering Applications of Artificial Intelligence 26: 779–788.
http://dx.doi.org/10.1016/j.engappai.2012.11.005
Wang, W. C.; Yu, W. D.; Yang, I. T.; Lin, C. C.; Lee, M. T.; Cheng, Y. Y. 2013. Applying the AHP to support
the best-value contractor selection – lessons learned from two case studies in Taiwan, Journal of
Civil Engineering and Management 19(1): 24–36. http://dx.doi.org/10.3846/13923730.2012.734851
Yager, R. 1981. A new methodology for ordinal multiobjective decisions based on fuzy sets, Decision
Science 12: 589–600. http://dx.doi.org/10.1111/j.1540-5915.1981.tb00111.x
234 S. Ulubeyli, A. Kazaz. Fuzzy multi-criteria decision making model for subcontractor ...

Yu, D. J. 2013. Intuitionistic fuzzy prioritized operators and their application in multi-criteria group
decision making, Technological and Economic Development of Economy 19(1): 1–21.
http://dx.doi.org/10.3846/20294913.2012.762951
Zadeh, L. A. 1965. Fuzzy sets, Information and Control 8(3): 338–353.
http://dx.doi.org/10.1016/S0019-9958(65)90241-X
Zadeh, L. A. 1973. Outline of a new approach to the analysis of complex systems and decision processes,
IEEE Transactions on Systems, Man, and Cybernetics 3: 28–44.
http://dx.doi.org/10.1109/TSMC.1973.5408575
Zavadskas, E. K.; Kaklauskas, A.; Vilutiene, T. 2009. Multicriteria evaluation of apartment blocks main-
tenance contractors: Lithuanian case study, International Journal of Strategic Property Management
13(4): 319–338. http://dx.doi.org/10.3846/1648-715X.2009.13.319-338
Zavadskas, E. K.; Turskis, Z.; Tamosaitiene, J. 2008. Contractor selection of construction in a competitive
environment, Journal of Business Economics and Management 9(3): 181–187.
http://dx.doi.org/10.3846/1611-1699.2008.9.181-187
Zavadskas, E. K.; Vilutiene, T.; Turskis, Z.; Tamosaitiene, J. 2010. Contractor selection for construction
works by applying SAW-G and TOPSIS grey techniques, Journal of Business Economics and Manage-
ment 11(1): 34–55. http://dx.doi.org/10.3846/jbem.2010.03.

Serdar ULUBEYLI is an Associate Professor of Construction Management Division in the Department

of Civil Engineering at Bulent Ecevit University, Turkey. His areas of academic research interests include
project management, subcontracting, and international construction projects. He has published many
papers in various scientific journals and proceedings.

Aynur KAZAZ is an Professor of Construction Management Division in the Department of Civil Engi-
neering at Akdeniz University, Turkey. Her areas of academic expertise include construction management,
total quality management, and contract management. She is the author of many papers published in
professional journals and conference proceedings.