Identifying Factors Affecting Dropout in Higher Education

Abstract

Identifying factors that influence academic dropout is crucial for coordinators to take preventive actions against student desertion. This thesis aims to explore the current state of research in the field, seeking to understand the main factors associated with academic dropout and how they can be applied in predictive models of dropout. A quantitative, exploratory, and explanatory applied research was conducted, using case study techniques and bibliographic research. The systematic literature review indicated that dropout could be predicted by academic, demographic, and learning factors. Using these factors, predictive models were developed with the help of machine learning algorithms, whose predictions proved sensitive to changes in course flows. It concludes that academic, demographic, and learning factors are effective in predicting dropout, with the selection and application method of these factors being crucial for achieving accurate predictive results. Moreover, the application of academic factors in predictive models requires consideration of the specific context of the analyzed data.