Confronting the growing disparity between standardized evaluation systems and personalized competency development in practical education, this study proposes a novel data-driven framework integrating Multi-Criteria Group Decision Making (MCGDM) to enhance curriculum assessment within Problem-Based Learning (PBL) environments. Specifically, the framework incorporates Z-number theory to effectively capture the uncertainty and reliability inherent in expert evaluations, addressing the challenges posed by subjective and imprecise human judgments. The well-established Multi-Attributive Border Approximation Area Comparison (MABAC) method is extended through the integration of Z-number modeling to enhance robustness in ranking and decision-making processes. A multi-stage assessment process is designed, encompassing a Pass check phase, a Score determination phase, and a Grading phase, aligned with the pedagogical principles of PBL. Furthermore, a hybrid Entropy-Criteria Importance Through Intercriteria Correlation (CRITIC) weighting scheme under Z-number representation is introduced to objectively determine the importance of evaluation criteria, considering both information dispersion and inter-criteria correlation. The proposed method is applied to a real-life case study involving 24 students, evaluated by multiple stakeholder groups, including peer teams, instructors, and industry experts. Sensitivity and comparative analyses suggest that the proposed framework provides a robust and reliability-aware assessment procedure within the studied course context. The findings indicate its potential to support a more transparent and structured evaluation process, although broader generalizability requires further validation across cohorts, courses, and institutions.
Yu et al. (Mon,) studied this question.