Uma proposta de análise de desempenho de estudantes do ENEM

Abstract

In the international education field, it is relevant to predict the students' performance for future courses and/or disciplines. In that matter, Rajalaxmi et al. (2019), Abledu (2012) and Gadhavi, Patel (2017) works show a statistic-mathematical approach by multiple regression which generates predictives models that may help educational politics, professors and managers in decision making. Starting from Educational Data Science, this Undergraduate Thesis uses real data from the Exame Nacional do Ensino Médio (ENEM). Starting from a sample of 26,731 participants, it shows that it is possible to model the student performance, to then correlate multidimensionally the four knowledge pillars of the ENEM, and, on top of that, provide a discussion about the students' performance from public and private schools. Results show that Educational Data Science is a fundamental field to Mathematics professors' formation, elucidating, from the quantitative point of view, the educational variables.