Investigation of variables explaining TIMSS 2019 mathematics achievement with multilevel structural equation model

In this study, it is aimed to examine the variables at the student and school level that are expected to affect the mathematics achievement scores of the eighth grade students who participated in the TIMSS 2019 Turkey assessment, using two-level structural equation modeling. In the study, mathematics achievement scores in the TIMSS 2019 evaluation, student and school questionnaire were used as data collection tools. While the variables of liking mathematics, self-confidence in mathematics, clarity of teaching in mathematics and the value given to mathematics were examined at the student level, the variables of mathematics teaching resources and the school's emphasis on academic achievement were used at the school level. Firstly, separate measurement models were created for both levels, and then a two-level structural model was established. While there was a significant negative correlation between success in mathematics and liking mathematics, the strongest predictors of success were found to be self-confidence in mathematics and the school's emphasis on academic achievement. In addition, a high positive correlation was found between self-confidence in mathematics and liking mathematics. The value given to mathematics and mathematics teaching resources did not show a significant relationship with success.