Empirical analysis of workers' perceptions of ESG impacts on corporate sustainability performance: A methodological innovation combining the PLS-SEM, PROMETHEE-ROC and FIMIX-PLS methods

Abstract

The perceptions of workers regarding Environmental, Social, and Governance (ESG) criteria are not only reflective of their immediate work environment but also serve as indicators of broader corporate sustainability performance. This research provides empirical insights into workers' perceptions of ESG impacts on corporate sustainability performance in the Brazilian electrical industry. It demonstrates the methodological strengths of Partial Least Squares Structural Equation Modeling (PLS-SEM), combined with the Preference Ranking Organization METHod for Enrichment Evaluations with the Rank-Order Centroid (PROMETHEE-ROC) and Finite Mixture Partial Least Squares (FIMIX-PLS) methods. A structured questionnaire administered across various sectors of two large companies captured comprehensive data on workers' ESG views. Cluster Analysis (CA) and Factor Analysis (FA) grouped and validated the data, while PLS-SEM assessed associations between latent and observable variables. FIMIX-PLS analyzed sample segmentation, and PROMETHEE-ROC ranked significant ESG criteria. Findings revealed that ESG criteria significantly influence workers' perceptions of corporate sustainability, with multidimensional FA elucidating 71 % of latent traits and PLS-SEM parameters demonstrating strong model fit. This study contributes to theory and practice by validating the PLS-SEM structure for capturing worker perceptions and by introducing a novel methodology combining PLS-SEM, PROMETHEE-ROC, and FIMIX-PLS methods.