Use of statistical modeling and data science to improve learning and reduce school inequalities in public schools

• Grant number: 22/06522-8
• Support Opportunities: Research Grants - Research in Public Policies
• Duration: February 01, 2023 - January 31, 2025
• Field of knowledge: Humanities - Education - Educational Planning and Evaluation
• Principal Investigator: Mozart Neves Ramos
• Grantee: Mozart Neves Ramos
• Host Institution: Instituto de Estudos Avançados (IEA). Universidade de São Paulo (USP). São Paulo , SP, Brazil
• Associated researchers: Alexandre Pereira Salgado Junior ; Antonio José da Costa Filho ; Carla Aparecida Arena Ventura ; Filomena Siqueira e Silva ; João Bosco Paraiso da Silva ; Juliana Chiaretti Novi ; Juliana da Silva Dias ; Leonardo Guimarães Medeiros ; Luiz Alberto Frezzatti Negreiros ; Marco Antonio Alves de Souza Junior ; Perla Calil Pongeluppe Wadhy Rebehy ; Rafael Naime Ruggiero ; Yago Silveira Marinzeck Santos
• Associated scholarship(s): 23/08737-4 - A principal component analysis (PCA) of educational indicators for the initial and final years of elementary education in public school systems, BP.TT

Abstract

To place quantity and quality in the same equation is certainly the greatest challenge of Brazilian public education. In the last decades, the country improved access to school, especially at the elementary level. However, Brazil must focus on its low learning outcomes and how to enhance them considering all the students and the extensive education inequalities among schools from the same education network, or among municipalities from the same states, or even among the states from the same Brazilian region. The state of São Paulo is not different from other states in Brazil, although it presents learning outcomes above the national average. Nevertheless, the state is still struggling to respond to the great educational demands of the 21st century, even if we consider municipalities with high socioeconomic status. This project aims at building statistical modelling for data analysis in order to improve learning outcomes and decrease inequalities among public elementary schools in the first and final years at the municipalities of Ribeirão Preto, Batatais, Cordeirópolis, Francisco Morato and Jundiaí, benefiting more than 100,000 students. Therefore, the multivariate statistical technique of Principal Components Analysis (PCA) will be used to establish patterns of school performance at the education public networks of these municipalities, enabling, in a second stage, the construction of an indicator of educational inequality for these schools in 2015, 2017 and 2019. In addition, the project also aims at investigating how school principals influence learning outcomes, using data from the SAEB´s socioeconomic survey, with questions related to leadership at the schools, based on the literature and on the students grades/scores at SAEB. Data will be statistically analyzed using multiple linear regression. Evidence shows that an effective leadership at the schools can impact learning outcomes in up to 12 points at SAEB´s exam. (AU)