A novel integrated objective-subjective weighting method using principal component analysis to correct overrepresentation of correlated criteria in sustainability decision-making

Indicators can be used as criteria in multicriteria decision-making. However, when indicators measure the same phenomenon, they exhibit correlations that can bias decision-making. Methods that consider correlations have limitations and are either objective or subjective. This paper proposes a new integrated objective-subjective weighting method that accounts for criterion correlations via principal component analysis. This method favours objective weights when criteria are correlated and subjective weights when they are uncorrelated. The method was validated through three comparative case studies and six sensitivity analyses, with a qualitative analysis of decision-relevant rank changes. Results show decision-relevant rank changes across all case studies, highlighting a correlated structure in decision-making in 11 of 13 MCDM models. 10 MCDM models report changes in the countries that constitute the top or bottom 5. 9 MCDM models report that most countries changed ranks, while 8 MCDM models report that countries have moved by more than 5 ranks. The sensitivity analysis highlights the method's contribution: it may enable more balanced decision-making. Pairwise comparisons allow for a more complete configuration that better captures the correlations among the criteria and their weights. The sensitivity analysis also identified limitations, including incomplete consideration of variance and computational complexity. Further research should apply principal component analysis to remove redundant criteria and validate the proposed method's decision quality.