Extended evaluation based on deviation from median attribute values (EDMAV) method for fuzzy multi-criteria decision-making

The classical Multi-criteria Decision-making (MCDM) methods such as the Evaluation based on Deviation from Median Attribute Values (EDMAV) assume that the evaluation values are crisp numbers. The EDMAV method combines the results of two different models to get the global score of each alternative, namely the weighted arithmetic mean (WAM) and weighted median (WMed). This combination provides a nuanced picture of central tendency that sheds light on the overall distribution and the typical values of data. Although, uncertainty, vagueness, and fuzziness are the main feelings that a person always confronts before making a responsible decision. Thus, this paper aims to extend the EDMAV approach to deal with the MCDM issues in a fuzzy setting. As an illustrative example, a case study of supplier selection was used to apply the suggested method. The ranking results obtained by the Fuzzy EDMAV method (A1 ≻ A5 ≻ A2 ≻ A4 ≻ A3) demonstrate a significant agreement with those obtained by the Fuzzy Evaluation based on Distance from Average Solution (Fuzzy EDAS) method which is evidenced by very high rank correlation coefficients ( \(r_s\) $r_s$  = 1). It was suggested to do a sensitivity analysis to assess the results' consistency and validity. After the inclusion of fourteen hypothetical weight settings, the generated ranking order showed a significant agreement proved by more than 0.8 as a value of rank correlation coefficients. This study's findings indicate that the extended EDMAV approach is reliable and effective.