From Survey to Success: Predicting First-Year Retention Through Longitudinal Data Analysis

• In Event: Data-Driven Insights into Student Retention, Attainment, and Beyond

Abstract

This study examines how a New Student Survey (NSS) can help predict first-year retention among full-time undergraduates at a large public research university. Using separate logistic regression models, we analyzed institutional and student-reported data to identify key predictors of dropout. While traditional metrics (e.g., SAT scores, academic warnings) explained more variance overall, several survey-based constructs (i.e., college expectations, learning attitude, and student well-being) emerged as significant predictors. This study demonstrates that new student surveys, when properly designed and validated, offer unique, early insights into student success. By combining survey and institutional data, colleges can better identify at-risk students and implement timely, student-centered interventions that support retention and align with institutional goals.

Authors

• Ting Wang, Rutgers University

• Chinyere Ugwuanyi, Rutgers University - New Brunswick

• Eduardo Molina, Rutgers University