Coelho Junior, Francisco Antonio2024-02-282024-02-282018-02-26https://hdl.handle.net/20.500.14135/960The aim of this research was to test predictive relationships between selected context variables – family, student and school – on the academic performance of the subjects in Science, Mathematics and Reading, in the 2015 edition of Pisa, in Brazil. It also aims to describe possible tangible effects of the exam results and the context questionnaires for the creation of public policies in education in Brazil, since it is expected that the results of Pisa may have subsidized governmental actions or public policies in education, in Brazil. The research is divided into two stages, one quantitative and the other qualitative. For the quantitative stage, exploratory/descriptive analyzes, correlation analysis and multiple linear regression analysis were performed. For the qualitative part, interviews with the representatives of the education secretariats were carried out, with the specific objective of describing the effects of information generated by Pisa in the process of formulating public policies in education. The results indicate that the Pisa information is not used or, at most, is little used. On the other hand, it was observed the full use of other internal (state) and/or external large-scale evaluations (such as SAEB / Prova Brasil). Most of the hypotheses of this study were confirmed to some degree, for some region of Brazil. It is concluded that state action is needed to foster greater use of the information that comes from Pisa for the formulation of more forceful actions to remedy part of the Brazilian educational problem. For future studies, it is suggested to investigate the use of other variables that may be impacting students' academic results and a detailed investigation of the practices of the States best placed in large-scale exams.Documento textualporAcesso abertoEducação ComparadaAvaliação da EducaçãoDesempenho AcadêmicoPolíticas Públicas em EducaçãoDissertação120808