Educational Data Mining in Higher Education: Building a Predictive Model for Retaining University Graduates as Master's Students

Abstract

The goal of this study was to create a model for predicting the factors that influence graduates’ decisions to continue their studies at the master's level within the same institution. The research was conducted on the entire population of students ( N = 663) who started their studies at the Faculty of Education, University of East Sarajevo between 2008 and 2018 and completed their studies by 2021. Part of the data was collected from the faculty information systems and part through questionnaires. The results showed the artificial neural network had the highest classification accuracy while variables, the personal factors, the faculty offers quality, applicable and useful study programs, time to degree and place of residence have the best predictive value. This model can enable other institutions of higher education to create an inclusive environment that enhances student wellbeing, improves educational results, and increases institutional efficiency.