<![CDATA[Achieving mathematics learning mastery at the elementary school level remains a significant challenge, requiring instructional approaches that support conceptual understanding and problem-solving abilities. This study aimed to estimate the pooled mastery proportion of elementary school students during Problem-Based Learning (PBL) implementation, examine between-study consistency, and evaluate the robustness of pooled estimates. A proportion meta-analysis was conducted on classroom action research (CAR) studies implementing PBL in elementary mathematics learning. Nineteen primary studies involving 448 students were included in the quantitative synthesis. Pooled mastery proportions were estimated using a random-effects meta-analysis of logit-transformed proportions. Between-study variance was estimated using Restricted Maximum Likelihood (REML), and 95% confidence intervals were calculated using the Hartung–Knapp adjustment. Estimates were subsequently back-transformed to the proportion scale. The results indicated that the pooled mastery proportion increased from 58.3% in Cycle I (95% CI: 53.0%–63.4%) to 85.5% in Cycle II (95% CI: 81.5%–88.7%), representing a descriptive increase of approximately 27.1 percentage points. Between-study heterogeneity was low in Cycle I (I² = 8.4%) and was not detected in Cycle II (I² = 0%). Leave-one-out sensitivity analysis confirmed the stability of pooled estimates after sequentially removing individual studies. Although all studies showed an increasing direction from Cycle I to Cycle II, the improvement should be interpreted as a change in mastery proportion during PBL implementation rather than evidence of a causal effect attributable solely to PBL, because the primary studies lacked control groups and involved repeated measurements of the same students. This study contributes evidence from a pooled mastery proportion perspective, which is more appropriate for synthesizing classroom action research outcomes than standardized effect-size approaches.]]>
Copyrights © 2026
