The effects of using of material to students' academic achievement in mathematics between 2013 and 2017: A meta-analysis

The purpose of the research is to examine the effect of material use to the academic achievement of students in mathematics teaching. As a result of the literature review, it was decided to limit the research to articles and thesis studies made between 2013-2017. A total of 24 studies, including 4 doctoral degree theses, 11 master's degree theses and 9 articles, were taken into account according to the inclusion criteria prepared in consideration of the research problem. In this study, meta-analysis was preferred as a method of the research because we wanted to find a statistically valid results at the end of the study. The reason for this is to obtain more objective results compared to the qualitative screening methods. It is also important that the research will meet the need for meta-analyzes in educational sciences. The meta-analysis study was done using metafor package in R statistic program. The effect sizes for each of the 24 studies, included in the study, were calculated using the statistical data given in the studies. The common effect size was calculated using random effects model, since the effect sizes for each of the 24 studies was found as heterogeneous as a result of the Q Cochrane heterogeneity test. Since we used random effects model, I2 statistic was used to assess the heterogeneity caused by variance. In this study, funnel plot, regression test and Kendalls' Tau test were used to assess the publication bias and finally the power of meta analysis was found using Fail-Safe N method. As a result of the meta-analysis, the common effect size of material use in mathematics teaching was calculated as 0.7564. According to common effect size of the research, the use of material has positive effect and 'large' influence on the achievement of student's in mathematics class. According to Cohen, the use of materials in teaching has 'medium' effect on student success, on the positive side. According to the p-value resulting from the Q test, the effect sizes of the studies are heterogeneous. For this reason, the random effect model was chosen and the common effect size was calculated. Since the research was not homogeneous, it was seperated into groups according to distinguishing characteristics for further investigation whether there was a significant difference between the groups. According to funnel graph, regression test and Kendall's Tau test no significant bias was found. Fail-Safe N value, indicating the power of the research, was found as 1491. This value for 24 studies shows that the meta-analysis is quite powerful.