preview

Increasing Student Retention : A Predictive Model Of Undergraduate Degree Non Completion

Better Essays

Increasing Student Retention: A Predictive Model of Undergraduate Degree Non-Completion

Abstract

This study seeks to develop a predictive model of college student dropout, using aggregate high school variables and individual postsecondary achievement variables to predict non-graduating students’ academic year of departure. After performing multiple linear regression and discriminant function analysis, the research found that a cohort of students admitted in the fall 2007 semester from several universities could be assigned an academic year of departure using data readily available by the end of a student’s third academic term. The university can use this model to predict student departure and improve the effectiveness of student retention efforts by focusing on targeted times when at-risk students are predicted to drop out.
Introduction
While many academic, psychological, and institutional variables influencing undergraduate student dropout have been studied, these factors have generally only been examined using models that treat student dropout as a binary, dependent variable. One of the shortcomings of using logistic regression in the study of undergraduate dropout is that it restricts the study’s ability to infer when a given student is likely to drop out. In contrast, the present study considered undergraduate dropout as occurring over a set of intervals, in this case academic terms, and sought to identify those crucial times when students are considering departure

Get Access