Overview of Analytic hierarchy process AHP
Analytic hierarchy process AHP is a common technique of multi-criteria decision making (MCDM), which contributes to investigating complicated issues of decision making. It is fulfilled by structuring the issue, identifying decision making factors, evaluating the importance of such factors, and the synthesis of all decision making factors (Saaty, 1980, 1982, 2008). Such an analytic decision making process is basically a way to measure intangible factors, which is shown by means of pair-wise comparisons and judgment on the dominance of an element over another one about a special attribute. This model is capable of examining both qualitative and quantitative elements (Saaty, 2005). One of the advantages of AHP is that it can change intangible factors into qualitative values and investigate the systematic evaluation of the weight of selected factors through a series of pair-wise comparisons (Saaty, 1980, 1982, 2008). Using pair-wise comparisons, which are applied based on experts’ judgments to compare the factors, let the decision maker compare just two factors and observe them freely, away from any irrelevant effects. Therefore, the foundation of AHP is the rational logic of binary comparison. Binary comparison creates relative importance values. Robert-Lowe and Sharp (1990) found out that planning an issue such as a hierarchy is a useful aid to understand the issues and to discuss them. This process can reveal the issues that have not already been explained clearly. Moreover, it is easy to understand such a process and decision makers can do comparisons very comfortably. This process is applied in four stages (Saaty, 1980):
۱. Designing a hierarchy that describes the issue.
۲. Planning matrices for binary comparisons between consecutive levels (classes).
۳. Presenting priorities or relative amounts of elements in each level of a hierarchy, which are obtained through Eigen values.
۴. Combining the relative amounts of different levels, which are obtained via the third stage so that a total score of alternatives is achieved.
Algebraic comparison of binary comparison is shown by Eq. (1).
In which aij reflects verbal judgments. If in Equation (1) the relationship is true for all comparisons, the matrix A is consistent (Saaty, 1980).
Relative priority vector W is calculated through the following formula:
The AHP method has been widely used in outsourcing models. (For instance, it has been used in studies conducted by Grewal, Saren and Gill, 2008; Yang, Kim, Nam and Min, 2006; Lee, 2011; Ho et al., 2009; Mahalik, 2011; Yang and Peng, 2012; Liu et al., 2008; Peng, 2012; Lai, 2012; Scott, Ho, and Dey, 2013). Of course, some studies on outsourcing logistics have applied the AHP model as well (for example, Grewal et al., 2008; Peng, 2012; Ho, Emrouznejad, He and Man Lee, 2012; Huang and Zhao, 2012; Shan, 2011). All mentioned cases have focused on outsourcing logistics services to 3PL in their research and none of them has evaluated important factors of decision making in the field of outsourcing logistics (whether forward or reverse). Moreover, a lot of studies have used a hybrid model of SWOT and AHP (Kahraman, Demirel and Demirel, 2007; Kurttila, Pesonen, Kangas and Kajanus, 2000; Masozera, Alavalapati, Jacobson and Shrestha, 2006; Shrestha, Alavalapati and Kalmbacher, 2004; Alshomrani and Qamar, 2012; Fabac and Zver, 2011; Oreski, 2012; Görener, Toker and Uluçay, 2012). This hybrid model is often used to improve the capability of using SWOT analysis because AHP can quantitatively determine the importance of the SWOT group factors (Kurttila et al., 2000). Görener et al. (2012) have stated that a lack of ranking the importance of factors in SWOT has led to inefficiency of this analytic model. Therefore, they combine SWOT analysis with the analytic hierarchy process so that quantitative strategic planning will be improved. Fabac et al. (2011) identified the significance coefficients of SWOT factors by means of the AHP model so that the quantitative analytic model would be improved. A brief review of previous studies shows that the inefficiency of the SWOT analytic model in quantifying the factors has caused this analytic strategic model to be combined with Analytic hierarchy process AHP technique. Accordingly, Shrestha et al. (2004) investigated agricultural prospects and challenges through the SWOT analytic model combined with AHP. Masozeraet al. (2006) combined AHP and SWOT to assess the estimated beneficiaries with regard to suitable community-based management approaches. Kahraman et al. (2007) used the SWOT and Analytic hierarchy process AHP models to prioritize the strengths, weaknesses, opportunities, and threats of the group in the first stage and to evaluate alternatives to electronic government strategy and make decisions about it in the second stage. Using the hybrid model of SWOT and AHP, Alshomrani et al. (2012) have recently identified key factors in strategic challenges for establishing electronic government in Saudi Arabia. They have concluded that user-centered strategy and digital gap elimination strategy are the best strategies available to create a successful electronic government. Oreski (2011) made use of the SWOT and Analytic hierarchy process AHP hybrid in strategic tourism planning for the city of Varazdin. Each one of these studies has used the case study approach to examine the validity of the studied method.
On the other hand, it has been proven that AHP is one of the most practical MCDM methods (Zahedi, 1968). Nevertheless, this technique has been widely criticized. Perhaps the most challenging issue of AHP is the change of “ranking,” which was immediately noticed after the cited method was introduced. Most of the research conducted by MCDM researchers has discussed ranking change criterion (for example, Belton and Gera, 1983; Dyer, 1990; Holder, 1991). The research conducted by Harker and Vargas(1987) is highly valuable and noticeable in validating different aspects of AHP. In most of new books, Saaty (2010) has drawn the AHP researchers’ attention towards remarkable characteristics and sensitive points of Analytic hierarchy process AHP. He believes that ranking changes, which follow structural changes, greatly contribute to using relative scales and lubricating structures. According to Tversky et al., (1990) the main issue is not keeping the rank; it is deeply believed that the rank should not and cannot always be kept. What matters is whether the “independence” hypothesis is considered or not. In general, the three principles of AHP are supported by four general rules, which are classified as follows: reversibility, comparison of homogeneous elements, system and hierarchical dependency, and hopes (expectations).These are associated with elements such as ranking credit, result value, and dependence of levels of activities on the structure that is used and their vastness (Saaty, 2010). Moreover, in the research conducted by Peniwati (2007), several groups of decision making methods were evaluated according to 16 pre-determined criteria in order to identify their usefulness and effectiveness. The results showed that,in comparison to other methods that were compared via relevant criteria, the Analytic hierarchy process Analytic hierarchy process AHP and ANP models are of great functionality. Accordingly, it could be found out that the AHP model is one of the most powerful MCDM methods and that is why it has been used in this research.
Since the SWOT analysis has some quantitative restrictions and identifying the relationship between the factors usually depends on the decision maker’s information and familiarity with the studied company, individuals’ judgments about the studied factors can influence the ultimate results (Kumar et al., 2009). Moreover, in spite ofthe successful and widespread use of AHP in lots of decision making issues, one criticism always remains: its inability to manage uncertainty due to assigning integers to decision makers’ perception (Deng, 1990). The natural way to deal with uncertain judgments or decisions is the use of fuzzy sets (or fuzzy numbers) in expressing comparative ratios. Due to further development of intuitionistic fuzzy sets rather than just fuzzy sets (Zadeh, 1965), the intuitionistic fuzzy analytic hierarchy process (IF-Analytic hierarchy process AHP) was combined with SWOT analysis in this research decision making model in order to examine strategic factors of outsourcing reverse logistics. The IF-AHP model was used for the first time in this study in the field of evaluating outsourcing reverse logistics. Despite the advantages of using the hybrid model of SWOT and Analytic hierarchy process AHP in conducted studies on outsourcing issues, this hybrid model has not been used so far in the field of outsourcing reverse logistics to analyze the strengths and weaknesses of the decisions. (Analytic hierarchy process AHP)
فرآیند تحلیل سلسله مراتبی – Analytic hierarchy process AHP