Investigation of variables explaining PISA 2018 reading literacy using multilevel structural equation model

Today, it is important to carry out studies on the Programme for International Student Assessment (PISA), which has a global impact. In this direction, it is aimed to determine the variables explaining the reading literacy from the answers given by the Turkish students to the PISA 2018 questionnaires, with the Multilevel Structural Equation Model, which also considers the sample weight and plausible values. Remaining 5295 students and 182 schools were included in the analysis by removing the missing values and the values found outside the analysis assumptions. As a result of the analysis, it was seen that 56% of the changes in students' reading literacy were caused by differences between schools and 44% by individuals. At student level, economic, social, and cultural status, expected occupational status, emotional support of parents, enjoyment of reading, interest in ICT, disciplinary climate and metacognition explain students' reading literacy, but mastery goals and ICT resources were insufficient to explain. At school level, while explaining the shortage of educational material and the behaviors of students and teachers that hindering learning; the variables of school type and student/teacher ratio were insufficient to explain reading literacy. In the model, which was retested with 12 variables in total, it was observed that the coefficient of determination was .228 at the student level and .327 at the school level, for a total of .555. Accordingly, 55.5% of the reading literacy of Turkish students participating in the PISA 2018 assessment can be explained by the twelve variables selected in this study.