A fuzzy hybrid approach for project manager selection

Article history: Received October 25, 2015 Received in revised format: December 12, 2015 Accepted January 2, 2016 Available online January 4 2016 Suitable project manager has a significant impact on successful accomplishment of the project. Managers should possess such skills in order to effectively cope with the competition. In this respect, selecting managers based on their skills can lead to a competitive advantage towards the achievement of organizational goals. selection of the suitable project manager can be viewed as a multi-criteria decision making (MCDM) problem and an extensive evaluation of criteria, such as Technical skills, experience skills, Personal qualities and the related criteria must be considered in the selection process of project manager. The fuzzy set theory and MCDM methods appears as an essential tools to provide a decision framework that incorporates imprecise judgments and multi criteria nature of project manager selection process inherent in this process. This paper proposes the joint use of the Fuzzy DEMATEL (FDEMATEL) and Fuzzy VIKOR methods for the decision-making process of selecting the most suitable managers for projects. First, with the opinions of the senior managers based on project management competency model (ICB-IPMA), all the criteria required for the selection are gathered. Then the FDEMATEL method is used to prioritize the importance of various criteria and FVIKOR used to rank the alternatives in a preferred order to select the best project managers from a number of alternatives. Next, a real case study used to illustrate the process of the proposed method. Finally, some conclusions are discussed at the end of this study. Growing Science Ltd. All rights reserved. 6 © 201


Introduction
Project is a temporary endeavor undertaken to create a unique product, service or result (Archibald & Archibald, 2013, p. 3).Nowadays, with the increasing complexity of organizations, requiring more complex and extended projects to meet their needs.Project managers are responsible for the leadership role in projects (Mϋller & Turner, 2010).The selection of well-trained managers ready for the leadership of the projects is a critical success factor for organizations.On the other hand employing a poorly prepared project manager, without the necessary skill, knowledge and experience, is something that could threat the success of the project (Pinto & Slevin, 1987;Mϋller & Turner, 2007).Successful project managers should have relevant experience and knowledge of the technology required by the project they manage (Zavadskas et al., 2012;Vainiunas et al., 2010).There are some traditional methods used in the project manager selection process such as, completion of application forms, interview and background investigation.These traditional techniques generally come to a conclusion on the basis of the subjective judgment of decision makers, which makes the accuracy of the results questionable (Zhang & Liu, 2011;Robertson & Smith, 2001;Petrovic-Lazerevic, 2001).Thus, In order to find the suitable project manager, developing effective selection methods is vital (Dagdeviren, 2010).Selection of the suitable project manager can be viewed as a multi-criteria decision making (MCDM) problem and an extensive evaluation of criteria, such as Technical skills, experience skills, Personal qualities and the related criteria must be considered in the selection process of project manager.As in many decision problems, project manager selection problem is extremely complex in real life; humans generally fail to make a good prediction for quantitative problems, in contrast they may make accurate guesses in qualitative forecasting.Because of the imprecise expressions, a fuzzy approach is commonly used in decision problems.Fuzzy sets theory appears as an important tool to provide a decision framework that incorporates imprecise judgments inherent in the project manager selection process and permit the translation of linguistic expressions into numerical ones.(Dursun & Karsak, 2010).This paper proposes fuzzy integrated multi-stages evaluation model towards enhancing the execution procedures of Project manager selection processes.The proposed methodology consists of fuzzy DEMATEL and fuzzy VIKOR methods under multiple criteria.First, with the opinions of the senior managers based on project manager's competency model (IPMA, 2006), all the criteria required for the selection are gathered.Then The FDEMATEL method is used to prioritize the importance of various criteria and FVIKOR used to rank the alternatives in a preferred order to select the best project managers from a number of alternatives.A real case study in Zarrin Balan Shomal Company is used to illustrate the process and potential benefits of the proposed framework.The rest of the paper is organized as follows: In the next section, comments on the recent literature are summarized as concerns project manager selection problem.A hybrid FMCDM model combining FDEMATEL and FVIKOR for assessment of project managers is developed in Section 3. In Section 4, an empirical case conduced in Zarrinbalan company, is presented to demonstrate the proposed method.Finally, conclusions and future research directions are provided in Section 5.

Literature Review
Multi criteria decision making (MCDM) has been used in selecting project manager.For example, Chen and Cheng (2005) developed a fuzzy MCDM method for information system project manager selection.Xing and A-di (2006) analyzed the significance of choosing an eligible project manager in their study.They tried to quantitatively assess the ability and quality of a project manager by implementing fuzzy analytical hierarchy process which was based on triangular fuzzy numbers.Hui et al. (2008) tried to demonstrate a suitable competency based framework.Zavadskas et al. (2008) developed a multi criteria methodology for project manager selection based on grey criteria.Liqin et al. (2009) adopted a fuzzy comprehensive evaluation methods in the selection of a project manager.Rashidi et al. (2010) combined fuzzy systems, ANNs and Genetic algorithm for choosing a qualified project manager.Insight into the relevant literatures to project manager's selection reveals that majority of the reviewed studies do not use combination of FDEMATEL and FVIKOR for project managers selection problem, which this is the novelty of this research.
The important step in project manager selection problem is to identify the selection criteria.Researchers have their own opinion on the project managers' selection criteria (El-Sabaa, 2001).Technical skills, conceptual skills and human skills are considered by Godwin (1983), as the main four skills project managers cannot do without.Analysis and design, Programming, Interpersonal skills, business knowledge, IS environment knowledge, IS applications knowledge are considered by Chen and Cheng (2005) as the main four skills project managers.Shih et al. (2007) considered Knowledge tests (language test, professional test, safety rule test), skill tests (professional skills, computer skills) and interviews as the main criteria for On-line manager Recruitment in a local chemical company.(Chen & Hung, 2012;Kelemenis et al., 2011;Chen & Lee, 2007;Afshari et al., 2012;Sadeghi et al., 2014;Xing & A-di, 2006;Torfi & Rashidi, 2011;Al-Harbi, 2001).Proper criteria selection is the building block for successful project manager selection.Hence this paper aims to provide a systematic method for criteria selection based on expert's idea and project manager's competency model.Table 1 shows extensive literature Review of papers relate to project manager selection problem.

A hybrid FMCDM model for evaluating project managers
In this section, a hybrid MCDM model based on FDEMATEL technique and fuzzy VIKOR method is presented to address the problem of project manager selection with interdependence among criteria.In short, the proposed model for evaluation of project managers consists of two main stages: (1) constructing the influential relation map (IRM) among the dimensions and criteria and calculating their influential weights by FDEMATEL technique, and (2) ranking project managers through the fuzzy VIKOR method.The flowchart of the proposed hybrid MCDM model is shown in Fig. 1.

The FDEMATEL Technique
The DEMATEL technique originated from the Geneva Research Centre of the Battelle Memorial Institute is especially pragmatic to visualize the structure of complicated causal relationships and to clarify the essentials of the problems (Baykasoğlu et al., 2013).It is more suitable in real-life applications and has been widely used in various decision making problems (Dalalah et al., 2011;Cebi, 2013;Liu et al., 2014).The DEMATEL technique can help decision makers understand the interdependence of criteria through matrices or digraphs and restrict the relations that reflect characteristics within an essential systemic and developmental trend.The procedure of the FDEMATEL algorithm can be given as follows: Step 1. Compute the group direct-influence matrix Let , , … , be a set of criteria, where denotes the jth criteria, j= 1, 2,…, n; , , … , be a set of experts, where represents the kth expert, K= 1, 2,…, K. First, experts are asked to indicate the direct effect that factor has on criteria , by applying linguistic variables (Lin & Wu, 2004).These linguistic terms are shown in Table 2.

Table 2
The correspondence of linguistic terms and linguistic values (Lin & Wu, 2004)
Step 2. Calculate the normalized direct-influence matrix by normalizing initial direct-relation fuzzy matrix, we acquire normalized direct-relation fuzzy matrix by using: ) 2 ( = , , " , " , " In Which: ) 3 ( Step 3. Derive the total-influence matrix Total-relation fuzzy matrix is defined as: Step 4. Defuzification of Total-relation fuzzy matrix by using: Step 5. Build the influential relation map (IRM) At this step, the sum of the rows and the sum of the columns within the total-influence matrix T are respectively expressed as the vectors and using: ) 9 ( ) 10 ( Where denotes the sum of the ith row in matrix T and shows the sum of the direct and indirect effects that factor i has on the other factors.Similarly, denotes the sum of the jth column in matrix T and shows the sum of direct and indirect effects that factor j has received from the other factors.Let i=j and , ∈ 1,2, . ., ; the horizontal axis vector (D+R) is then defined by adding r to c, Which illustrates the strength of influences that are given and received of the factor.That is, (D+R) shows the degree of the central role that the factor plays in the system.Similarly, the vertical axis vector (D-R) is created by subtracting R from D, which shows the net effect that the factor contribute to the system.If is positive, then factor j has a net influence on other factors, and if is negative, then factor j is being influenced by other factors on the whole.Finally, an IRM can be acquired by mapping the ordered pairs of (D+R, D-R), which provides more valuable information for problem solving.
Step 6. Determine the influential weights of criteria After the DEMATEL confirms the influential relationships between the dimensions and criteria, we use the causal diagram to measure the criteria weights that will be used in the decision making process (Dalalah et al., 2011;Baykasoğlu et al., 2013).The relative importance of the criteria is calculated with the following equation: ) 11 ( The weight of any criterion can be normalized as follows:

Where
represents the final criteria weights to be used in the decision making process.

The FVIKOR technique
VIKOR is one of the multi-criterion decision making techniques, which was developed by Opricovic (1998) and Opricovic & Tzeng (2002) with Serbian name; VlseKriterijumska Optimizacija I Kompromisno Resenje, means multi-criterion optimization and compromise solution (Pal & Gauri, 2010).The VIKOR was developed as a multi criteria decision making technique to solve a discrete decision problem with non-commensurable and conflicting criteria (Opricovic & Tzeng, 2004;Sanayei et al., 2010;Opricovic & Tzeng, 2007).Fuzzy VIKOR intends to find the decision-makers preferable compromise that suits human objective cognition (Wang, 2006).It is generally hard to find an alternative that meets all the criteria simultaneously, so a good compromise solution is preferred.This problem may become more complex when multiple decision-makers with different perception on the alternatives are involved.Thus, VIKOR is a useful technique for ranking and sorting a set of alternatives.There is uncertainty related to the research data.Then fuzzy MADM was used.Fuzzy logic suggested that decision makers can use linguistic variables.Fuzzy VIKOR was described By Wang and Chang (2005).The procedure of the FDEMATEL algorithm can be given as follows (Opricovic 2011;Opricovic & Tzeng, 2007;Shemshadi et al., 2011;Wang, 2006;Huang et al., 2009): Step 1. Form a group of decision-makers (denoted in n), then determine the evaluation criteria (denoted in k) and feasible alternatives (denoted in m).
Step 2. Define linguistic variables and their corresponding triangular fuzzy numbers.Linguistic variables are used to evaluate the importance of the criteria and the ratings of alternatives with respect to various criteria.(As shown in Table 3).

Table 3
Linguistic variables for rating the alternatives (Sun, 2010) Linguistic terms Linguistic value Very poor (0, 1, 3) Poor (1, 3, 5) Moderate (3, 5, 7) Good (5, 7, 9) Very good (7, 9, 10) Step 3. Construct a fuzzy decision matrix.Calculate fuzzy weighted average and construct the (normalized) fuzzy decision matrix, where is the rating of alternative with respect to criterion , and is the important weight of the jth criterion.This study, therefore, denotes linguistic variables and as triangular fuzzy numbers.
Step 4. Determine the fuzzy best value (FBV) and fuzzy worst value (FWV): Step 5. Calculate the values: where refers to the separation measure of from the fuzzy best value, similarly, is the separation measure of from the FWV, and is the weight of each criterion.
Step 6. Calculate the values of: The index min is with a maximum majority rule, and min is with a minimum individual regret of an opponent strategy.As well, ν is introduced as weight of the strategy of the maximum group utility, usually ν = 0.5, whereas 1-ν is the weight of the individual regret.
Step 7. The next step is defuzzification of triangular fuzzy number , and to crisp number.Here the defuzzification method, the second weighted mean, is applied to convert a fuzzy number into crisp score, which is shown as next Equation ) 26 ( B 2 4 Step 8. Rank the alternatives, sorting by the values , and in ascending order. Step 9. Propose as a compromise solution the alternative which is best ranked by the measure (minimum) if the following two conditions are satisfied.
C .Acceptable advantage: where is the alternative with second position in the ranking list by: ) 28 ( Q: DQ .Acceptable stability in decision-making: The alternative must also be best ranked by or/and .This compromise solution is stable within a decision making process, which could be the strategy of maximum group utility (when ν > 0.5 is needed), or "by consensus" ν ≈ 0.5, or with vector (ν ≤ 0.5).Here, ν is the weight of decision making strategy of maximum group utility.If one of the conditions is not satisfied, a set of compromise solutions is proposed, which consists of:  Alternatives and if only the condition is not satisfied, or  Alternatives , , …, if the condition C is not satisfied; is determined by the relation: For maximum M the positions of these alternatives are in closeness.

Case study: selection of project manager
In this section, an empirical study conducted in an Iranian company (zarrin balan shomal) which is the best producer of Livestock and poultry food in mazandran, is presented to illustrate the application of the proposed hybrid decision making model.In this company Stakeholders decided to build a large office building in Babol city.They needed a good manager for this project.Therefore 6 experts and (0.00,0.09,0.31)(0.00,0.08,0.26)(0.00,0.00,0.16)(0.07,0.17,0.45)(0.00,0.04,0.24)C1 (0.00,0.07,0.26)(0.00,0.02,0.2) (0.00,0.00,0.14)(0.00,0.03,0.22)(0.00,0.09,0.3)C 2 (-0.01,0.26,0.54)(0.06,0.16,0.43)(0.00,0.00,0.17)(-0.12,0.4,0.7)(-0.11,0.34,0.64)C3 (0.11,0.16,0.43)(0.00,0.03,0.22)(0.00,0.02,0.22)(0.28,0.29,0.62)(0.24,0.25,0.56)C4 (-1.11,0.04,0.22)(0.00,0.08,0.28)(0.00,0.00,0.18)(-2.76,0.25,0.55)(-2.41,0.18,0.45)C5  which shows the importance of all criteria by aggregation of all managers' preferences and which assign criteria into cause and effect groups.As the shown in Table 9, the criteria are divided into two groups.The first is cause group which implies C1, C3, C4 and effect group which implies the rest of the criteria.Then, the IRM of the given application can be constructed by using the vectors d and r in the total-influence matrix T, as expressed in Fig. 2. Finally The relative importance of the criteria calculated by Eq. ( 12).The results are depicted in Table 10.The weights of the alternatives are calculated by FDEMATEL up to now, and then these values can be used in FVIKOR.So, the FVIKOR methodology must be started at the second step.According to the Table 3, linguistic terms converted and aggregated to the triangular fuzzy number and decision matrix was constructed as Table 11.Then, weighted normalized decision matrix can be prepared.This matrix can be seen from Table 12.By following FVIKOR procedure steps and calculations, the ranking of project managers are obtained.The results and final ranking are shown in Table 13.According to the above empirical study, the proposed hybrid FMCDM model provides some important findings.First, in accordance with the results of FDEMATEL (see Table 10), Personal qualities (C4) is the most important criterion with influence weight of 0.247849, and Technical level (C2) is the least important one with influence weight of 0.144943.Second, the FDEMATEL method can also be utilized to understand the interrelationship among dimensions and criteria (Fig. 2).The IRM shows that Level of leadership (C3), Personal qualities (C4) and Site management capacity (C1) criteria have more influence over the other two criteria.This finding means that they are the most important relative to the other criteria.Third, from the results obtained by fuzzy VIKOR (see Table 13), the ranking order of the project managers alternatives is A2>A1>A3>A5>A4, suggesting (A2) as the most suitable project manager for this project.

Conclusions
The selection of a project manager from a set of potential candidates is an important, difficult, and timeconsuming task for managers of any company.This problem worsens with an increase in the number of alternatives.There is also a risk of human error in judgment and decision making.There is therefore a need for computational models that can increase the accuracy of decisions and reduce the time required.An appropriate and simple prioritization method for determining the best project manager would be helpful to firms.A two-step fuzzy-DEMATEL and Fuzzy-VIKOR methodology is structured here that FVIKOR uses FDEMATEL result weights as input weights.Then a real case study in zarrin balan shomal firm presented to show applicability and performance of the methodology.First, with the opinions of the senior managers based on project management competency model (ICB-IPMA), all the criteria and required for project manager selection are gathered.Then The FDEMATEL method is used to prioritize the importance of various criteria and FVIKOR used to rank the alternatives in a preferred order to select the best project managers from a number of alternatives.The results of this study showed that the most important criteria for project manager selection was Personal qualities (C4), and the most suitable project manager was alternative (2).As a future step to this paper could be the comparison of the proposed approach to other FMCDM methods, like FTOPSIS, FAHP, FANP or even more to the outranking methods, such as FELECTRE and FPROMETHEE.

Table 1
Baykasoglu et al. (2007)ent applicationsBaykasoglu et al. (2007)considered Communication skills, technical expertise, problem solving ability, decision making skills, available time period, salary request as the required skills for Project team member's selection.Insight into the relevant literatures to project manager selection reveals that majority of the reviewed studies do not provide a systematic method for criteria selection Education, Planning and control, Communication, Experience, Leadership, Negotiation skills, General management, Team development, Resource management, Time management, Human skills, Technical skills, Computer, Problem solving, Knowledge (Technical competences, Behavioral competences, Contextual competences), Experience(Technical competences, Behavioural competences, Contextual competences)Hadad et al. (2013) Mann-Whitney-Wilcoxon U test project manager selection Criteria regarding the project's allocated budget and actual Costs, project's resources and their consumption, the project's time span, project's risks Chen et al. (2013) FTOPSIS Project Leader Selection General ability: English proficiency, communication ability, work attributes and emotional steadiness professional ability: French language ability, product knowledge, management capacity, Risk management capability, Ability to create efficient working environment), Personal qualities(Achievement-oriented and good at action, Ability to influence, Cognitive ability), Contextual competences(Project orientation, Program orientation, Portfolio orientation, Project program and portfolio implementation, Permanent organization, Health, security, safety and environment)

Table 9
Casual relationships between criteria

Table 10
Relative weights of criteria