Predicting Students’ Eligibility for Higher Education from Teachers’ Leadership Behaviors and Students’ Perceived Stress

Little evidence exists on the factors that enable students in upper secondary school to achieve eligibility for higher education. This lack of knowledge limits educators’ and policymakers’ opportunities to identify students who may require additional educational support before graduation. In this study, we apply machine learning to predict students’ eligibility for higher education using survey data from 272 Swedish students in upper secondary school. Our results show that students’ eligibility for higher education can be predicted with an average accuracy of 85.5%. Program type emerged as the strongest predictor, while teacher leadership behaviors and students’ perceived stress also contributed to the prediction model. Among the teacher leadership behaviors, performance-oriented and supportive leadership showed the highest predictive importance. In addition, students’ perceived stress related to the risk of not completing their studies emerged as an important predictor of eligibility. With this study, we extend previous research on predicting educational outcomes by focusing on students’ eligibility for higher education and by integrating theoretically grounded measures of teacher leadership behavior and students’ perceived stress into prediction models. Furthermore, we contribute to the growing discussion on the role of teacher leadership behavior and students’ perceived stress in predicting students’ progression towards higher education and demonstrate the potential of machine learning to support the early identification of students who may require additional educational support.