TOPSIS-RAD Fixes Rank Reversal by Anchoring to Decision-Maker Goals
June 8, 2026
TOPSIS-RAD replaces dataset-derived ideal solutions with two decision-maker-defined reference arrays: Vetoed Performance Levels exclude non-viable alternatives before normalization, and Desired Performance Levels cap scores at aspirational thresholds. This eliminates rank reversal sensitivity to outliers and anchors rankings to explicit requirements rather than dataset extremes.
HOW THIS AFFECTS YOU
●
researcherWorth watching for multi-criteria decision analysis work where rank reversal and outlier sensitivity are known failure modes in TOPSIS-based evaluation pipelines.