Minerando dados para entender o impacto da pandemia da COVID-19 no Exame Nacional do Ensino Médio

Abstract

In Brazil, the main performance evaluation exam in basic education is the National Secondary Education Examination (ENEM), which is also used to enroll students in higher education. In 2020, with the arrival of COVID-19, basic education institutions needed to change their educational model from face-to-face teaching to the use of distance learning methodologies, which may have affected the quality of education received. Therefore, there is a need to understand the effects that the COVID-19 pandemic caused to ENEM. This master’s thesis aims to identify the main impacts caused by the pandemic on ENEM, considering all of Brazil. To this end, the complete process of Educational Data Mining (MDE) was carried out, based on the Cross Industry Standard Process for Data Mining (CRISP-DM) methodology, in ENEM in 5 different years, in order to understand the years pre-pandemic and the first two years of the pandemic. In particular, this study used techniques of descriptive exploratory analysis, correlation and hierarchical grouping to identify the impacts caused on the ENEM. The results show that participants with higher incomes performed better, private schools outperformed other types of schools, the number of present and absentees in the exam was lower in the pandemic years, and there was a change in the socioeconomic characteristics of the participants. . Finally, the pandemic did not negatively impact student performance, but the characteristics of the participants who took the exam changed and the number of absentees and present in the exam was drastically impacted.