Ciência de dados para análise da desigualdade social durante a pandemia: uma análise com dados públicos brasileiros

Abstract

The data analysis process has been widely used in research across various fields of knowledge, including demographic studies and investigations into social inequality during the COVID-19 pandemic. Among the common approaches, those based on public data made available by government agencies stand out. In Brazil, relevant information on education, economy, health, and sociodemographic aspects is accessible to the general public. In this context, this study employs data science techniques to analyze Brazilian sociodemographic and educational data, aiming to measure inequality relations during the COVID-19 pandemic and provide technical support for public policy formulation. The research was guided by two main hypotheses: (1) inequality and social vulnerability play a crucial role in the worsening of epidemic outbreaks; and (2) lower-income social classes are the most afiected by such crises due to limited access to quality basic services, such as healthcare and education. Based on these hypotheses, two case studies were conducted using public data. The results of the first study indicate that students from public schools in situations of socioeconomic vulnerability are more likely to miss the ENEM (National High School Exam) in scenarios where the exam is institutionally suspended, such as during states of public emergency. The second study found that low-income areas with poor urban infrastructure are more susceptible to increases in confirmed COVID-19 cases and related deaths. Additionally, the number of residents per household was shown to be associated with higher transmission rates of respiratory diseases, reinforcing the role of social vulnerability as a factor that amplifies the efects of the pandemic. The first case study analyzes the efects of school closures on student participation in the ENEM during the pandemic, from 2020 to 2022. The second assesses the impact of educational, urban, and economic infrastructure conditions in the Brazilian population using the Social Vulnerability Index (SVI), seeking to identify the factors most strongly associated with increases in COVID-19 cases and deaths.