Quaestum

Multilevel analysis applied to performance factors in Paraná's Public Education (2017-2021)

Gláucia Marília Hass — 2025-05-29

This study presents the results of a research developed as an MBA's in Data Science & Analytics final paper and it aims to analyze IDEB's performance factors in the public system of education from the province of Paraná, through information analysis obtained from federal databases, from external evaluations and SAEB's censitary data. The methodology of analysis and data treatment was done through supervised technique of Machine Learning called Multilevel Hierarchy Modelling.

Hierarchical sectoral grouping of Brazilian industry with input-output data

Danilo Cristian da Silva Sousa — 2025-04-15

This study proposed a sectoral analytical methodology for the Brazilian economy, using the input-output analysis, originally developed by Leontief for the United States. Through the analysis of groups and networks, based on data from the 2015 Input-Output Matrix (IOM) and the 2019 Input-Output Tables (IOT) of the Brazilian economy, sectors were grouped according to performance variables in foreign trade, labor income generation and labor productivity. This grouping made it possible to identify a positive differential in certain industrial sectors compared to the agricultural and service sectors, placing them in a more advanced position in Brazil's path towards the income of developed countries. Furthermore, through network analysis, it was observed that the main consumers of the production of these industrial sectors are external demand and the households’ consumption, indicating that, if the country seeks development through the expansion of industrial production, it is crucial consider alternatives for the sustained growth of industrial exports and the consumption of certain industrial goods by families.

Análise de cluster da evasão escolar em escolas públicas de Goiás: segmentação e implicações para políticas educacionais

Wesley Silva Ferreira — 2025-05-14

The present work addresses the problem of dropout in public schools in the state of Goiás, using cluster analysis to identify patterns and guide educational policies. The relevance of the topic lies in understanding and reducing dropout rates, which has a significant impact on the educational system. The general objective was to identify patterns of school dropout in the public education network in the state of Goiás through cluster analysis. Cluster analysis techniques were used to group municipalities in Goiás with similar characteristics in relation to evasion. Dropout rates from 1st to 9th grade and high school in 246 municipalities were analyzed in the period 2019 and 2020. The results highlighted the complexity of dropout rates at different stages, with emphasis on the final and initial years of elementary school. from highschool. The analyzes revealed variations in dropout rates. Furthermore, some clusters presented rates below the national average. The practical implications are evident, with recommendations for targeted interventions in clusters 2 and 3, as well as in the identified outliers. In conclusion, cluster analysis proved to be an effective tool for understanding and addressing school dropout. The results obtained provide support to guide more effective, personalized and contextualized educational policies and strategies in future work. Furthermore, this study can serve as a basis for future research that wishes to delve deeper into this topic.

Startup Social: o programa USP na Escolinha como caso de empreendedorismo e impacto social

Delaine Goulart da Rocha — 2025-03-18

Startups emerge in different scenarios with a modelable proposal and, in some cases, in situations of uncertainty, which is why they come under the scrutiny of entrepreneurs. The prospecting of the business model in companies is positive and the structuring is confirmed by pointing out its characteristics that are clearly identified in existing ventures. In this sense, pointing out elements in companies consolidates the identities of startups and favors the progression of the model. This paper seeks to analyze the "USP na Escolinha" Program as a social startup. The “USP na Escolinha” Program was created in 2013 and is an extension program of the Faculty of Animal Science and Food Engineering that takes advantage of the formative years of elementary school children to introduce concepts and inspire desire and curiosity about the courses offered by the Pirassununga Campus. In order to achieve this objective, a quantitative and qualitative methodology was adopted, which allowed us to analyze the sum of the components that made up the proposal. The quantitative analysis is based on the documents recorded between 2013 and 2022, which constitute the development of the program. The qualitative analysis consists of interviews and personal statements made during the period, which are multiple-choice questionnaires collected on the perception of participating users. The results obtained show the identification of elements that present themselves in the startups and from there support the creation of repeatable standard models for the same idea.

Exploring the correlation between sanitation standards and school performance in Brazil: a clustering analysis

Karoline Louize Macedo Granja — 2025-08-01

Sanitation and education are two critical areas requiring immediate attention and improvement in Brazil. By increasing investments in these sectors, we can enhance the population's overall well-being through improved education and health outcomes. When students face health issues, it becomes challenging for them to succeed academically, as they may experience hospitalizations, difficulties concentrating due to illness, and even absences from classes. Although various factors can contribute to illness, this paper specifically focuses on diseases resulting from inadequate sanitation conditions. The primary objective of this research is to examine the correlation between sanitation standards and academic performance indices, considering sanitation, education, and health indicators. This research explores both positive and negative correlations, going deeper into the states’ clustering based on these indexes and their correlations. The data collected in this study are sourced from prominent government databases, namely the Brazilian Institute of Geography and Statistics [IBGE], Anísio Teixeira National Institute for Educational Study and Research [INEP], and the Unified Health System [SUS], covering the period from 2010 to 2020. Results indicate a clear link between inadequate sanitation standards and higher rates of hospitalization, or low sanitation rates and poorer school performance indexes.

Efeitos das unidades federativas na renda disponível per capita por domicílio: uma análise multinível

Eduardo Ben Hur de Queiroz Gomes — 2025-05-22

Este trabalho investigou os efeitos das unidades federativas brasileiras na renda per capita por domicílio, utilizando microdados da Pesquisa de Orçamentos Familiares (POF) 2017-2018, realizada pelo Instituto Brasileiro de Geografia e Estatística (IBGE). O objetivo foi analisar como a média de anos de estudo e a condição de segurança alimentar dos domicílios influenciam a renda disponível familiar per capita, considerando a forma como as unidades federativas podem impactar essa renda. Foram utilizados os microdados da POF aplicando regressão multinível, adequada para capturar os efeitos hierárquicos utilizando as variáveis explicativas: média de anos de estudos e condição de segurança alimentar. A escolha dessas variáveis foi sustentada pela literatura que indicou a educação como fator crucial na melhoria da renda familiar e na promoção do desenvolvimento econômico, e a segurança alimentar como indicador direto das condições de vida e saúde, influenciando a qualidade de vida e a capacidade de gerar renda. Os resultados mostraram diferenças significativas na renda entre domicílios de diferentes unidades federativas. A média de anos de estudo dos moradores e a condição de segurança alimentar impactaram significativamente a renda familiar per capita. Concluiu-se que as políticas públicas devem considerar essas variações regionais para promover distribuição de renda mais equitativa.