Study of novel normalization technique on weighting and ranking methodology in multi criteria decision making

The evaluation of corporate financial performance typically entails the use of multiple financial ratios characterized by heterogeneous properties, including the potential presence of negative values. Within this context, Multi-Criteria Decision-Making (MCDM) approaches have been extensively employed due to their capacity to systematically integrate diverse evaluation criteria into a coherent and objective ranking framework. However, most conventional normalization techniques in MCDM are not specifically designed to accommodate negative data, thereby potentially resulting in distorted weights and inconsistent ranking outcomes. This study proposes a novel normalization approach, termed the Root Distance Ratio (RDR), which is derived from the Weitendorf principle and formulated through a non-linear root transformation. The effectiveness of the RDR method is evaluated by integrating it into four MCDM techniques PSI, ARAS, MABAC, and EDAS across three corporate financial data scenarios, namely datasets with entirely positive values, negative values within benefit criteria, and negative values within cost criteria. The results demonstrate that, across the evaluated cases, the RDR method produces normalized values within the [0,1] interval, generates more proportionally distributed weights, and yields more consistent ranking outcomes compared to conventional normalization techniques. These findings indicate that RDR enhances both the stability and interpretability of financial data-driven multi-criteria decision-making processes, particularly in the presence of negative values.