An Analysis of Data Production Based on the Consistency of Decision Matrices

Abstract

Multi-criteria decision making methods are used to solve numerous problems related to several disciplines such as engineering, management and business. Consistency of a decision making application is of crucial importance because of its dependability, reliability and sustainability. In this study, data production phenomenon is discussed based on consistency of a decision matrix. We complete different tests including (i) 16 tests (different triangular scales) for different 64, 125 and 216 decision matrix sets for four-criteria problems, (ii) 16 tests for different 256, 625 and 1296 decision matrix sets for five-criteria problems. Based on our observations from the results, we find that first level derivation is always consistent. However, the second and higher level derivations exhibit inconsistencies. The results are also valid if the expert evaluations for the same pairwise comparisons are considered as equal (1, 1, 1). This study is expected to improve the reliability results for the future decision making studies.