Investigating the measurement invariance of the school belonging and student bullying scales in PIRLS 2021

The purpose of the research is to investigate the measurement invariance of the Sense of School Belonging and Student Bullying Scales according to gender and culture (Turkey, England, and Hong Kong) within the framework of PIRLS 2021 Progress in International Reading Literacy Study. The sample for the study consisted of 4th-grade students participating in PIRLS 2021 study from Turkey, England, and Hong Kong. The data collection instruments for the study consisted of the Sense of School Belonging Scale and Student Bullying Scales. In this context, the results of the Principle Component Analysis (PCA) for the scales and reliability evidence in terms of Cronbach's Alpha coefficients for each country were presented. The single-factor structures of the scales were confirmed by using Confirmatory Factor Analysis (CFA). Following the confirmation of the scales, a measurement invariance study was conducted using Multiple Group Confirmatory Factor Analysis (MGCFA) techniques within the framework of Structural Equation Modeling (SEM). According to the findings, while configural, metric, and scalar invariance were provided for the sense of school belonging scale and student-bullying scale across gender subgroups, strict invariance was not provided. In the cultural analysis, it was found that only configural invariance was provided for both scales, while metric, scalar, and strict invariance were not provided. As measurement invariance was provided by gender, it can be concluded that all comparisons made according to gender would be meaningful. However, as measurement invariance was not provided across cultures, it can be concluded that comparisons of scores obtained from the scales would not be meaningful for Turkey, England, and Hong Kong.