This study aims to determine the differences in the effectiveness of the Problem Solving and Problem Based Learning models on elementary school students' mathematical critical thinking skills using Meta-analysis. The first step in this research is to determine the problem. The second step is searching for research journal data through Google Scholar; after completing the data search, the researchers obtained 20 appropriate journals. The third step is to analyze the research data to get solid results and conclusions, and the fourth or final step is to report the research results by conveying the essential parts of the research results obtained. The comparison of the Problem Solving learning model showed an average increase of 14.64%, while the Problem Based Learning model was 28.73%. The prerequisite test in this study uses the normality test, homogeneity test, and linear test. A normality test using the Shapiro-Wilk technique shows that the significance value is > 0.05. Pretest homogeneity test showed Sig. of 0.314 > 0.05 while the posttest showed Sig. of 0.676> 0.05. The significance value in the linearity test of the two learning models is > 0.05. ANCOVA test using Univariate shows that the value of Sig. 0.002 means less than 0.05 (0.002 <0.05). The results of the ANCOVA test prove that if f count < f table is 0.003 < 3.69 and the significance value is 0.960 > 0.05, which proves that Ho and Ha are accepted. So, the conclusions can be drawn from the results obtained to prove if there are significant differences. Furthermore, the Effect Size test shows that the Problem Solving and Problem Based Learning models have a relatively great influence on thinking critically in mathematics.
                        
                        
                        
                        
                            
                                Copyrights © 2021