Passed to Fail? Predicting the College Enrollment of GED® Passers

Abstract

Utilizing a dataset of over 900,000 enrollees in adult education programs in New York State between 2005 and 2013, we examine college-going of GED® passers. Upon enrollment, participants were asked whether they wanted to attend college after completion; almost 13,000 students both indicated a desire to attend college and subsequently passed the GED® Exam. Roughly half of these students reported attending college within 12 months. We use logistic regression to predict which students attended college based on a number of demographic variables and proximity to a community college. Counter to prior research, we find that none of these variables are practically significant predictors of college attendance. Students of different races who are and are not employed, receiving public assistance, single parents, and living close to a community college (among other factors) are virtually equally likely to attend college. We discuss other possible explanations and recommend that future research examine noncognitive and other factors.

Keywords: GED