A Systematic Review and Meta-Analysis of Fuzzy Logic for Students’ Performance Assessment

Abstract

Assessment of students’ performance using fuzzy logic provides high flexibility and reliability in education. This study aims to apply a systematic review and meta-analysis to synthesize the findings of selected studies in this field, with a focus on the effects of fuzzy logic system configurations on students’ performance assessment outcomes, focusing on subgroup analyses to explore variations across input and output factors. A total of 109 articles were retrieved from databases, including ScienceDirect, Springer, IEEE, Scopus, and Google Scholar, for qualitative and quantitative syntheses. Among these, 46 studies reported both fuzzy and non-fuzzy median scores and were included in the meta-analysis. Results showed that output member functions achieved the highest median scores (≥ 89.50), and subgroup analyses revealed significant heterogeneity across studies (I² ≥ 75%, p <0.01). Frequency-based combinations of fuzzy variables generally outperformed non-frequency configurations, enhancing system granularity and accuracy. These findings highlight the importance of optimizing fuzzy logic system designs to improve student performance assessment.