The effect of instruction with mathematical modeling on mathematics achievement and attitude: A meta-analysis study

The aim of this research is to determine the effect of teaching with mathematical modeling on mathematics achievement and attitude by using meta-analysis method. At the same time, the studies were examined with the moderators of publication year, publication type, country of application, research design, teaching level, learning area/subject, duration of application and sample size. Studies were determined by using different combinations of mathematical modeling, achievement and attitude keywords and English equivalents in Higher Education Institution (YÖK) National Thesis Center, ULAKBİM-DergiPark Akademik, ProQuest Dissertations & Theses, Web of Science (WOS), Scopus and ERIC databases and Google Scholar search engine. According to the inclusion criteria, 51 effect sizes were obtained from 45 studies on mathematics achievement and 21 effect sizes from 19 studies on mathematics attitude. The data obtained from the studies were analyzed with the CMA (Comprehensive Meta-Analysis) program. Hedges's g value was used for the effect sizes of the studies and the mean effect size was calculated according to the random effects model. Funnel plot, Rosenthal's safe N, Orwin's safe N, Egger's linear regression analysis, Begg and Mazumdar rank correlation, Duval and Tweedie's trim and fill methods were used to test whether there is publication bias in the research. According to the findings of the study, it was determined that there was no publication bias in the studies included in the meta-analysis and there was sufficient heterogeneity among the studies. As a result of the research, the mean effect size of teaching with mathematical modeling on mathematics achievement was found to be high with 0.845 (%95 CI, SE=0.068) in the confidence interval of 0.712-0.978, the mean effect size of teaching with mathematical modeling on mathematics attitude was found to be medium with 0.420 (%95 CI, SE=0.083) in the confidence interval of 0.257-0.582. While the effect of teaching with mathematical modeling on mathematics achievement differed significantly according to the year of publication, country of application, research design, learning area/subject and sample size moderators, it did not differ according to the moderators of publication type, teaching level and duration of application. While the effect of teaching with mathematical modeling on mathematics attitude differed significantly according to publication year, publication type, teaching level and learning area/subject moderators, it did not differ according to other moderators. At the end of the study, some suggestions were made to practitioners and researchers about teaching with mathematical modeling.