Relative Pairwise Ranking: A normalization-free ordinal ranking approach for multi-criteria decision-making

Abstract

In Multi-Criteria Decision-Making (MCDM) methods, selecting an appropriate data normalization technique is a complex process. An unsuitable normalization method may distort the intrinsic ranking of alternatives. To address this issue, this study proposes a normalizationfree, ordinal pairwise ranking approach, referred to as Relative Pairwise Ranking (RPR). The method evaluates alternatives based on criterion-wise pairwise dominance and aggregates these outcomes using criterion weights, providing a scale-independent and interpretable ranking framework. The performance of the proposed approach is evaluated through illustrative examples and a real-world case study, supported by sensitivity and comparative analyses. The results indicate that RPR produces stable and consistent rankings across different weighting schemes and demonstrates robustness in the tested rank-reversal scenarios. These findings suggest that the proposed approach offers a practical alternative to normalizationbased MCDM methods.