Using Predictive Modeling to Increase One-Year Retention Through Early Identification

Abstract

One-year retention is a critical benchmark for student success and an important metric for university administrators. Initiated by the Retention & Graduation Task Force at a large, public, 4-year university with a plateaued one-year retention rate, this study generated a predictive model that can be used to identify students at risk of not being retained. After exploring many models to identify the most significant predictors of retention at three different time-points during a student’s first year, the researchers calculated predictive probabilities to assign students a risk level. The researchers shared this information in a way that could be actionable to advisors and others in the campus community. Evaluation is currently ongoing to determine if these initiatives for early intervention have a positive effect on retention.

Date
Mar 1, 2022 4:00 PM — 5:00 PM
Event
Texas Association for Institutional Research 44th Annual Conference
Location
Denton, TX
Avatar
Jorge Martinez
Director, Data Science

Director of Data Science at Houston Independent School District.