| Susan Rotermund
University of California, Santa Barbara
Can schools reduce dropout rates by improving student engagement? Using multilevel structural equation modeling to explore how schools affect students' dropout decisions
FINAL REPORT: The Role of Psychological Precursors and Student Engagement in a Process Model of High School Dropout
Drawing on data from the Education Longitudinal Study 2002 (ELS:2002), this study examined psychological factors, student engagement and student achievement in relationship to the process of dropping out. Student engagement has been included in many theoretical models of the dropout process, but few empirical studies have investigated the role of engagement in dropping out. Even fewer studies have investigated the theory that certain psychological factors related to motivation are important antecedents to engagement, and that consideration of these factors is essential to understanding the dropout process. This study used Exploratory and Confirmatory Factor Analysis to test a proposed measurement model of psychological and engagement constructs, including three engagement factors: affective, cognitive, and behavioral; and three psychological factors: sense of belonging, valuing school, and perceived competence. Global fit indices and strong measurement indicator values confirmed the validity of the proposed scale. Structural Equation Modeling was then used to test a process model of dropout that suggested that psychological factors are important precursors to engagement, and engagement serves as a mediator between psychological factors and student outcomes of achievement and dropping out. Statistical results supported the viability of the tested model. Affective engagement was found to be predictive of cognitive and behavioral engagement, with behavioral engagement the only direct predictor of dropout. Results confirmed previous findings that student achievement has a direct effect on the risk of dropping out, with higher achievement levels associated with lower risk of dropping out.
This study has important implications for interventions designed to reduce dropout rates. It suggests that interventions should address students' psychological needs for competence, relatedness, and relevance of schooling before trying to improve student engagement as a means to keep more students in school. Since behavioral engagement is the most proximal predictor of dropout, measures targeted at improving behavioral engagement (including attendance and misbehavior) may occur too late in the dropout intervention process to be effective. Improving student-teacher relationships may be the more effective intervention, as it affects all aspects of students' psychological connections to school.
Back to Funded Dissertation Grants Page |