摘要：Problem-based learning (PBL) has been widely applied as a learning method in mathematics education. This meta-analysis study was conducted to summarize the evidence on the effectiveness of PBL on students' comprehension abilities. An associated database was searched for studies that met the inclusion criteria. The estimation method used a random-effect model with a 95% confidence interval (CI). The study analyzed 35 effect sizes from 22 primary studies published between 2011 and 2020, and for a total of 1993, students were included. The Comprehensive Meta-Analysis (CMA) program was used to assist the analysis. As a result of the study, the overall effect size was 0.790 (moderate effect) with a standard error of 0.181. Analysis of the study characteristics revealed that the application of PBL was more effective in the conditions of the sample size set by 1 - 30 students; lower level of education, and combined with the use of mathematics learning software. These findings become a consideration for mathematics education practitioners for the implementation of PBL in the future.
