Índice de evasão e tempo de permanência nos cursos de Ciências Atuariais e Estatística da UFS : uma abordagem estatística e de modelos de sobrevivência

Abstract

The Higher Education system in Brazil is characterized by an elitist structure, primarily designed to meet the specific demands of privileged groups while neglecting scientific production and compromising critical and inclusive education. In recent decades, public policies have played a crucial role in expanding and democratizing access to universities. Nevertheless, students still face numerous challenges that hinder course completion, leading to high dropout rates, particularly in science, technology, engineering, and mathematics programs. Given this context, this study investigates the factors influencing dropout rates and student retention in the Department of Statistics and Actuarial Sciences at the Federal University of Sergipe between 2013 and 2023. Using survival analysis methods, such as the Kaplan-Meier estimator and competing risks models, the study identified that the probability of dropout in the department is higher than that of course completion. Additionally, entry age and quota status were found to be significant variables influencing dropout rates. Subsequently, Cox regression models were fitted for both dropout and course completion events. In the dropout model, it was observed that non-quota students have a 30% higher risk of dropping out compared to quota students and that mixed-race students are 30% more likely to drop out than white students. Meanwhile, in the course completion model, only entry age remained significant, indicating that an increase of one year in a student’s entry age reduces their likelihood of completing the course by approximately 7%.