Meta-Analysis: The Effectiveness of Realistic Mathematics Education (RME) and Ethnomathematics-Based Learning on Students’ Mathematical Problem-Solving Abilities

Source abstract

This study aims to estimate the pooled effect size of Realistic Mathematics Education (RME) and ethnomathematics-based learning on students’ mathematical problem-solving abilities, as well as to analyze the heterogeneity and potential publication bias of the reviewed studies. This study employed a meta-analysis method, with the study identification, selection, and reporting process guided by PRISMA 2020. Articles were searched through Google Scholar and OpenRead, covering publications from 2020 to 2025. Based on the established inclusion criteria, 22 articles were included, consisting of 11 studies on RME and 11 studies on ethnomathematics-based learning. The included studies employed quantitative experimental or quasi-experimental designs and provided sufficient data on means, standard deviations, and sample sizes to calculate effect sizes. Effect sizes were calculated using Hedges’ g and synthesized using a random-effects model. The results showed that the pooled effect size for RME was 1.047, with a 95% confidence interval of 0.7194–1.375, while the pooled effect size for ethnomathematics-based learning was 0.9178, with a 95% confidence interval of 0.6699–1.166. Both effect sizes were categorized as very large. The heterogeneity analysis showed significant heterogeneity in the RME group (Q = 28.25; p = 0.002; I² = 65.27%), whereas the ethnomathematics-based learning group did not show statistically significant heterogeneity (Q = 17.04; p = 0.074; I² = 40.82%). Egger’s test indicated no statistically significant evidence of asymmetry in the RME group (p = 0.310), whereas the ethnomathematics-based learning group showed evidence of asymmetry (p = 0.034). Therefore, the pooled effect size of ethnomathematics-based learning should be interpreted with caution due to the indication of potential publication bias.