Risk Factors for Dropouts of University Students Assessed by Statistical Modeling

Abstract

Supporting university students retain their studies is important for both the students and universities involved from educational and university management perspectives. Herein, we applied statistical modeling to questionnaire data regarding the dropouts of students upon enrolling in a private university of social welfare. Moreover, we considered the mechanism by which the students suspended their studies. The model used herein simultaneously expressed common trends across departments and department-specific mechanisms related to dropouts; it also predicted the probability of dropouts based on the responses to the questionnaire. Results suggested that the factors leading to dropouts varied across the departments in accordance with its specialties and whether the students took a national examination or not. The results were validated by the interviews of two experts, and they confirmed that the outputs of the proposed model were reasonable, while there still remained some important issues which seemed to be worthy of consideration in our future investigations.