Factors Influencing High School Graduation and Dropout for At-Risk Students: A Multilevel Approach to Measure School Effectiveness

• In Event: Investigating Program Impacts Using Multilevel Models

Abstract

This paper predicts high school graduation and dropout for at-risk students in one of the largest school districts in the United States using 2007-2010 Florida high school graduation cohort. We predicted at-risk students’ graduation and dropout incorporating student and school level predictors employing a two-level hierarchical generalized linear model (HGLM). The study found significant effects on graduation and dropout due to academic and disciplinary factors (such as achievement, absenteeism, and suspensions) at student and school levels. The school level model predicting student dropout status produced a medium effect size of 0.45.

Authors

• Bidya Raj Subedi, School District of Palm Beach County

• Mark Howard, Palm Beach County School District

• Nancy Reese, School District of Palm Beach County, Florida

• Dione Christy, School District of Palm Beach County