Multi-level structural equation modelling for factors affecting science achievement of Korean middle school students before and after COVID-19: based on 2018–2020 national assessment of educational achievement data

Science academic achievement is a key construct for examining students' science learning outcomes. While international comparative assessments like TIMSS and PISA have shown high science achievement of Korean middle school students, there is a lack of comprehensive investigation of factors influencing the achievement using domestic data. Additionally, the impact of COVID-19 on science achievement-related factors remains largely unexplored. This study utilises multi-level structural equation modelling to examine the factors affecting science achievement among Korean middle school students. By analysing data from the National Assessment of Educational Achievement (NAEA), we identify individual- and school-level variables that affect student science achievement. The model fitting results were compared before (2018–2019; N = 5771 and N = 5614) and after (2020; N = 4,623) the COVID-19 outbreak. The analysis results reveal that individual-level factors, such as self-efficacy, affective domain of science learning, and self-studying time had significant effects on science achievement consistently throughout years. School-level factors, including principal's supervision, school climate, and science teacher's efficacy, demonstrated significantly positive effects on science achievement only in 2020 following the COVID-19 outbreak. This study enhances our understanding of the characteristics of high-achieving Korean middle school students and highlights the role of school community in supporting student learning during uncertain circumstances like the COVID-19 pandemic.