Kajian Teknik Normalisasi Root Distance Ratio pada Teknik Pembobotan dan Perangkingan MCDM: Beberapa Kasus dalam Kinerja Keuangan

Abstract

Multi-Criteria Decision Making (MCDM) is a discipline within the field of operations research that is frequently used for decision-making to determine the best alternative based on relevant criteria. In the case of ranking companies based on financial per formance, three cases are applied to represent the entire data characteristics that are suspected to produce irrational weights and rankings. All alternatives and criteria are formed into a decision matrix for each case. The construction of the decision ma trix is the most fundamental step in MCDM. A decision matrix that contains negative values can affect the normalization results, leading to inconsistencies in the resulting weights and rankings. Therefore, an appropriate normalization technique is needed to address this inconsistency. This study aims to evaluate a new normalization technique called Root Distance Ratio (RDR) for use in weighting and ranking within MCDM. The RDRnormalization technique is tested on several MCDM methods, namely PSI, ARAS, MABAC, and EDAS. PSI is a method that generates both weights and rankings, while the other methods are used for ranking purposes only. The results indicate that RDR is capable of identifying irrationalities in the rankings of negative data and produc ing more rational weights and rankings.This is also supported by the spearman rank correlation values between the hybrid methods, which range from 0.7 to 1.0.