Variance Measure in Decision-Making Process: An Interpretation

This paper considers variance measure that is popularly used in multi-attribute decision-making (MADM) either as a holistic approach or as a snippet of a holistic approach for determining values of key decision entities such as attributes and/or alternatives. Specifically, we attempt to understand the results that emanate from the variance measure through a detailed interpretive construct, which will serve as a guide to decision-making scholars for rational usage of the popular approach in their respective MADM problems. We present illustrative examples to clarify the interpretations. Furthermore, we explore how the interpretive construct can be extended to analyze consistency in attribute weighting, thereby strengthening the reliability of MADM results. We also demonstrate how the variance measure can complement rank-based and aggregation-based methods, offering a unified perspective that integrates statistical interpretation with decision-making logic.