A NOVEL APPROACH TO MEASURING CRITERION WEIGHTS IN MULTIPLE CRITERIA DECISION MAKING: CUBIC EFFECT-BASED MEASUREMENT

Abstract

In the realm of multi-criteria decision making (MCDM) literature, various approaches exist for quantifying the weight coefficients of criteria. In this study, unlike other methods of calculating weight coefficients, a mathematical model based on cubic interactions among criteria has been proposed (Cubic Effect-Based Measurement). This model aims to enrich the MCDM literature while providing a means to compute weight coefficients of criteria. The dataset employed in this investigation comprises criterion values extracted from the Global Innovation Index (GII) evaluations for 19 G20 countries. Through the analysis outcomes, the efficacy of the proposed methodology in objectively deriving criteria weight coefficients for different nations is demonstrated. Furthermore, a comparative analysis is conducted, juxtaposing the proposed method with other objective weighting techniques (ENTROPY, CRITIC, SD, SVP, LOPCOW, and MEREC) as part of a sensitivity analysis. According to the findings, it has been observed that the rankings of GII criteria measured by the CEBM method are distinct from those obtained through the application of other methods. Following the sensitivity analysis, a total of 10 scenarios were created by assigning varying quantities to the GII criteria of countries. Subsequently, the weights of GII criteria were ranked in comparison to both the CEBM method and other techniques. The results indicate that in each scenario, the rankings identified within the scope of the CEBM method differ from those determined by the alternative methods. Based on all of these findings, the sensitivity level of the CEBM method has been deemed to be high. According to another finding, it has been observed that the CEBM method exhibits a higher degree of similarity with the MEREC method in terms of discrimination distance and correlation analyses. Consequently, it is anticipated that the proposed methodology will make substantial contributions to both the domain of cubic functions and the wider MCDM literature.