Predictors of Dropout Status for Eighth- and Ninth-Grade Students in a Large School District

• In Event: 73.028 - Early-Warning Indicators and Indicators of Success

Abstract

A large school district in the southeastern United States has been investigating risk factors for students in danger of dropping out of school. To date we have explored two research questions: (1) which Grade 8 factors predict graduation status (for when students are expected to graduate)? and (2) which Grade 9 factors predict graduation status (for when students are expected to graduate)? The earliest cohort of students for whom we have sufficient longitudinal data is the entering eighth grade class of 2007-08. Specifically, logistic regression and discriminant analysis were compared in terms of predictive utility in predicting dropout status as of the 2011-12 school year. Predictors of dropout status differed slightly between 8th grade and 9th grade, but agreed overall.

Authors

• Jonathan Darrell Rollins, University of North Carolina - Greensboro

• Marty Ward, Winston-Salem - Forsyth County Schools