Enhancing supplier selection reliability: integrated variable returns to scale-robust parameter R approach

Effective decision-making requires a suitable supplier selection model that can consistently identify and recruit suppliers, especially in complex scenarios involving multiple variables. Recently, integrated-data envelopment analysis techniques, such as fuzzy-based methods, cross-efficiency, and cooperative games have demonstrated promising outcomes. However, these approaches have yet to address the challenge of accounting for variations caused by undetectable supplier indicators. To overcome this limitation, this study examines the relationships between input selection criteria and supplier indicators in a way that minimizes the impact of variations stemming from undetectable indicators on supplier performance responses. To achieve this, we propose an integrated and refined model that combines revamped variable returns to scale (VRS) with robust parameter estimation and multivariate multiple-dependent regressions. By incorporating the VRS methodology, our model ensures that the selection procesation technique also helps mitigate the influence of outliers and extreme observations, enhancing the model's reliability and robustns accounts for variations in scale efficiency across suppliers, enabling a fair and accurate comparison. The robust parameter estimess. Furthermore, the multivariate multiple-dependent regressions approach allows for considering multiple dependent variables simultaneously, enabling a comprehensive evaluation of supplier performance. This approach considers the interrelationships between performance indicators and ensures a more holistic assessment of supplier capabilities. In conclusion, our integrated refined model offers a comprehensive and effective solution for supplier selection in complex cases.