Regressão logística em microdados da educação

• Ano de defesa: 2023
• Autor(a) principal: Salvador Neto, Antônio
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal do Espírito Santo; BR; Mestrado Profissional em Matemática em Rede Nacional; Centro de Ciências Exatas; UFES; Programa de Pós-Graduação em Matemática em Rede Nacional
• Palavras-chave em Português: Desempenho acadêmico; Educação; Regressão logística; ENEM; Análise descritiva; Matemática
• Link de acesso: http://repositorio.ufes.br/handle/10/12615

Academic performance is a constant concern in the field of education. Statistical regression techniques applied to data obtained from the National High School Exam (ENEM) can assist in understanding the factors that influence students’ academic performance. The work provides theoretical foundations on linear and logistic regression, as well as discretization techniques, data manipulation, and model quality measures. Logistic regression models were fitted with ENEM data from the years 2020, 2021, and 2022 independently and for each assessment subject. The models proved robust enough to predict students’ performance based on socioeconomic data. Descriptive analysis and model coefficient examination point to a strong negative correlation between the number of people living in the same residence as the student and their performance. Additionally, the family income category, father’s occupation, and the type of school the student attended have a significant impact on performance in the assessment. Another important factor was age; higher ages tend to belong to higher performance categories.