Measurement Invariance of the Social, Academic, and Emotional Behavior Risk Screener - Teacher Rating Scale

Abstract

School-based universal screening in the social, emotional, and behavioral (SEB) domains allows for the early identification of students in need of SEB support. Importantly, equitable assessment in universal screening for the SEB domains is critical to engage in accurate and ethical data-based decision-making. Measurement invariance is one method for examining potential inequities in assessment tools, permitting the ability to evaluate which assessments or assessment items perform differently across groups. As such, this study utilized multi-group confirmatory factor analysis to evaluate the extent of measurement invariance for a commonly used universal screening tool for the SEB domains: the Social, Academic, and Emotional Behavior Risk Screener - Teacher Rating Scale (SAEBRS-TRS). The sample consisted of 1949 students in kindergarten through fourth grade in a Midwest, suburban school district. Examination of factor structures indicated the bifactor model yielded adequate fit and was utilized for measurement invariance testing. Multi-group confirmatory factor analysis results provided preliminary evidence that the SAEBRS-TRS displays invariance across a variety of student characteristics. Specifically, results supported configural and metric/scalar invariance of the bifactor model across the student characteristics of racial or ethnic identity, sex assigned at birth, and eligibility for free or reduced-price lunch. Yet, future research is needed to corroborate these findings. Limitations, implications for practice, and directions for future research are discussed.