Investigation of the factors affecting science achievement using TIMSS 2015 data

The main purpose of this study is to investigate the factors affecting 8th-grade students' science achievement in Turkey using TIMSS 2015 data. For this purpose, TIMSS 2015 data covering 6079 students at 218 schools in Turkey were analyzed by conducting structural equational modeling (SEM). First, the latent variables to be included in the SEM analysis were determined using an exploratory factor analysis (EFA). Then, a structural model was constructed to explore the student-level variables affecting science achievement. The results show that the variables of students' socioeconomic status (SES) and science self-efficacy, bullying, teachers' guidance in science teaching, and student value science explain 51% of the variance in students' science achievement. Furthermore, SES and science self-efficacy are found to be the most important variables to predict science achievement. However, there is no statistically significant relationship between student value science and science achievement. Consequently, there exists a statistically significant relationship between the variables of SES, science self-efficacy, bullying teachers' guidance in science teaching with students' science achievement.