Estimating Priorities in Group AHP Using Interval Comparison Matrices

In this paper analytic hierarchy process (AHP), a well-known approach for handling multi-criteria decision making problems, is discussed. It is based on pairwise comparisons. The methods for deriving the priority vectors from comparison matrices are examined. The existing methods for aggregating the individual comparison matrices into a group comparison matrix are revised. A method for aggregation, called WGMDEA, is proposed for application in the case study. Because exact (crisp) values cannot always express the subjectivity and the lack of information on the part of a decision maker, the interval judgments are more suitable in such cases. Two main methodological problems emerge when dealing with interval comparison matrices in group AHP: a) to aggregate individual crisp preferences into the joint interval matrix, b) to calculate the weights from the joint interval comparison matrix. In the paper we first discuss the already proposed approaches to the aggregation of individual matrices, and the derivation of weights from interval comparison matrices, pertained to AHP group decision making methodology. Then, a new method, ADEXTREME, for generating the interval group judgments from individual judgments is proposed. A numerical example based on Rural Development Program of the Republic of Slovenia in 2007-2013 is presented to illustrate the new methodology for deriving the weights from interval comparison matrices. The results obtained by WGMDEA, MEDINT and ADEXTREME methods are compared.