Identifying at Risk Students

The issue of undergraduate student retention has a major economic impact at both the university level and the individual student level. Retaining students costs less than acquiring new students. Furthermore, retention rates have a substantial impact on U.S. News & World Report magazine’s rank of academic excellence of U.S. national universities and liberal arts colleges. From the students’ point of view, educational achievement raises their lifetime earning potential. Because of these economic factors, much research has focused on improving retention rates.

This paper focuses on retention improvement efforts at a public liberal arts university. Critical to these efforts is the use of multivariate logistic regression models for identifying students at greatest risk for withdrawal prior to their second year of college and prior to their third year of college. The first-year retention model relies on pre-enrollment data so that “at risk” students can be identified for intervention services prior to the start of classes their first semester. The second-year retention model adds variables relevant to students’ first-year college experience. Intervention services include targeted advising and use of an academic support center.

Discussion will focus on construction of the models, identification of variables found to be significant, as well as both quantitative and qualitative results of the retention efforts.