This meta-analysis investigates the effectiveness of the Problem-Based Learning (PBL) model in enhancing students' critical thinking and problem-solving skills in science education. A total of 42 peer-reviewed articles published between 2020 and 2024 were analyzed, covering diverse educational levels and scientific disciplines. The analysis revealed a large average effect size (ES = 0.798; 95% CI: 0.782–0.813), indicating a strong positive impact of PBL. The low standard error (SE = 0.043–0.060) reflects high estimation precision. Statistical tests, including funnel plot symmetry, Trim-and-Fill analysis (p = 0.991), and Egger’s regression test (p = 0.947), confirm the absence of significant publication bias. Moreover, residual heterogeneity was minimal (τ² = 0.000), validating the appropriateness of the fixed-effects model. These findings provide robust empirical evidence that PBL is consistently effective across various contexts in improving students’ 21st-century competencies. The study offers meaningful implications for educators and policymakers in designing science learning strategies that foster higher-order thinking skills.
Copyrights © 2025