Issue: 2027/Vol.37/No.1, Pages 1
CAN CORRESPONDENCE ANALYSIS SERVE AS A RANKING METHOD? A COMPARISON OF THE CA-WLD APPROACH WITH MULTI-CRITERIA DECISION-MAKING TECHNIQUES
Ömür Demirer
This is not yet the definitive version of the paper. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article.
Cite as: Ö. Demirer. Can correspondence analysis serve as a ranking method? A Comparison of the CA-WLD approach with multi-criteria decision-making techniques. Operations Research and Decisions 2027: 37(1). DOI 10.37190/ord/241845
Abstract
This study evaluates correspondence analysis (CA) as a ranking tool for multi-criteria decision-making (MCDM). The proposed Correspondence Analysis--Weighted Load-Distance (CA-WLD) method converts alternative-criterion distances in the CA space into a weighted decision score. In a Monte Carlo simulation of 20,242 contingency tables, nine established MCDM techniques agree closely, whereas CA-WLD yields distinct information that depends on the interaction of weighting with data structure; a max-min baseline shows that this is not merely non-compensation. A controlled perturbation analysis shows that CA-WLD withholds first place from a classical leader whose lead rests on volume or on above-competitor concentration, and endorses it once that concentration falls or its peak criterion is weighted sufficiently. Applied to a real site-selection problem (241 respondents), CA-WLD diagnoses the popularity bias of count-based methods. It is a complementary diagnostic tool, not a rival.
Keywords: correspondence analysis, multi-criteria decision making, ranking, non-compensatory evaluation, Monte Carlo simulation
Received: 26 June 2026 Accepted: 16 October 2026
Published online: 16 October 2026