Abstract:
Mathematics scores differ between countries in international examinations. Factors related to this variation are investigated in various contexts, whereas curriculum related factors have not been considered as part of a comprehensive study until now. Therefore, this study was designed to examine possible factors contributing to variability between countries in Trends in International Mathematics and Science Study (TIMSS) 2015 mathematics test scores. In this correlational research, outcome of interest is TIMSS mathematics scores and predictor variables were extracted from TIMSS student, teacher and curriculum questionnaires. Factors were considered at three levels; country (curriculum), school (classroom context) and student. This study includes 219,788 students in 6,586 schools from 36 countries. Hierarchical linear modeling was selected for analyses due to its usefulness in applying group and individual predictors in their levels simultaneously. Findings show that mathematics scores varied significantly at each level, and total variance was decomposed as 41.1%, 22.8% and 36.0% respectively in student, school and country levels. Results show that countries with more centralized systems and more time assigned on curriculum for mathematics lessons performed lower. More home study support, less working load for teachers, more frequent homework assignments and slightly higher number of students in classes effected to achievement positively. Interactions show countries with more centralized systems, schools with high homework frequency and less number of students performed higher, whereas countries with less centralized systems could achieve higher with more home study support students.
