An Analysis of Predictive Factors for Student Attrition of Local Private University in Non-capital Region

In recent years, Korean higher education has been facing a dual crisis: the decline in the school-age population and the increase in student attrition, particularly threatening the sustainability of private universities located in non-metropolitan areas. This issue should not be regarded solely as a demographic problem but rather as one that requires active institutional efforts to reduce student attrition. Accordingly, this study aims to identify predictors of student attrition and propose solutions by analyzing K University, a private institution located outside the metropolitan area. For this purpose, data from the Korean National Survey of Student Engagement (K-NSSE) between 2020 and 2023 were utilized, and both Random Forest and logistic regression models were employed to extract the predictors of student attrition. The findings are as follows: First, higher levels of career maturity and university engagement were associated with lower probabilities of dropout. Second, paradoxically, higher levels of effective teaching strategies were associated with an increased probability of student attrition. Based on these results, this study suggests the necessity of implementing various extracurricular programs and counseling services to enhance students’ career maturity and university engagement in order to reduce student attrition rates. Furthermore, it emphasizes the importance of conducting continuous institutional analyses of attrition-related factors and shifting from an achievement-centered approach to providing programs tailored to individual student characteristics.