Sensitivity Analysis and Validation in MCDM Methods: A Comprehensive Review with Advancements, Applications, and Future Directions

The accelerated development of multi-criteria decision-making (MCDM) tools in engineering, management, environmental planning, and financial analytics has increased the demand for more precise and responsive weighting methods. Stereotypical subjective and objective weighting methods tend to be biased, inflexible, or fail to react to dynamic decision settings. The review summarizes a series of newly developed novel weighting approaches, such as hybrid, data-driven, machine-learning-assisted, and uncertainty-adaptive approaches, to address these limitations. Drawing from a broad scope of published literature, this paper assesses the methodological developments, including fuzzy-enhanced weighting, the use of entropy-based extensions, CRITIC variants, Bayesian and probabilistic schemes, combined Delphi-ANP-DEMATEL approaches, best-worst method (BWM) refinements, and new deep-learning-based weight estimators. The review points out the ability of such modern approaches to boost robustness, decrease subjectivity, and increase transparency in decision-making in a variety of MCDM models. It also establishes the gaps in methodology, current trends, and future research directions and sets new weighting strategies as a key tool in managing complex, uncertain, and high-dimensional decision problems. This combination helps them gain a better theoretical insight as well as provide some practical advice to researchers and practitioners wishing to apply advanced weighting methods in MCDM.