Risk Modeling of First Year Student Retention at a Community College: An Early Exploration of Data, Feasibility, and Application

Abstract

Set in a medium sized, suburban, two-year community college, this study explores the initial development of an actionable predictive risk model for first-year retention based on data currently collected and to identify target variables upon which data should be collected to improve future versions of the model. Institutional data were analyzed by means of Analysis of Variance and logistic regression and ordinary least squares regression analyses. Results suggested there are several important variables for which data as available but the risk model as developed has relatively low predictive power. The results of this study are used to inform College administrators, faculty, and staff about risk modeling, conclusions that can be drawn from existing data, and provide guidance on additional relevant variables upon which data should be collected. Additionally, recommendations for future research are discussed for this study’s institution and the field of education as it relates to Community Colleges.