A scale independent and distribution sensitive approach in multi-criteria decision-making (MCDM): The Canberra distance performance measurement (CADPM) method

This study introduces the Canberra Distance Performance Measurement (CADPM) method, a scale-independent and distribution-sensitive framework developed to address persistent methodological limitations in Multi-Criteria Decision-Making (MCDM), including linear dependency, scale heterogeneity, and inadequate responsiveness to micro-level variations. CADPM evaluates alternatives through proportional, component-wise comparisons and eliminates the need for predefined criterion weights, as the intrinsic structure of the Canberra metric neutralizes scale effects and preserves relative positional relationships. The study also incorporates the Proximity Index Value (PIV) method to ensure a comprehensive comparative setting and to examine weight-free reference-based evaluation under a unified analytical perspective. The empirical assessment comprises both real-data and simulation-based case studies designed to observe the method’s robustness under heterogeneous distributions. Sensitivity analyses demonstrate that CADPM is strongly resistant to ranking instability and remains unaffected by systematic weight perturbations. Comparative analyses reveal high concordance between CADPM and established techniques such as MARCOS, TOPSIS, CODAS, WEDBA, and PIV, confirming that the method exhibits structural compatibility while preserving its distinctive proportional sensitivity. Simulation experiments further show that CADPM maintains stable performance patterns across diverse scenario structures, underscoring its consistency under controlled variance, scale, and distributional shifts. The primary advantage of CADPM lies in its ability to amplify subtle proportional differences, particularly in small or near-zero values, which conventional absolute-difference metrics often suppress. This capability enables more equitable and discriminative evaluations in asymmetric or heterogeneous datasets. Overall, CADPM offers a theoretically coherent, computationally stable, and empirically validated contribution to MCDM research. Future work may explore parametric extensions, hybrid integrations with reference-based models, and broader applications across complex decision environments.