Exames curriculares e resultados educacionais: uma análise do exame nacional do ensino médio

• Ano de defesa: 2010
• Autor(a) principal: Camelo, Rafael de Sousa
• Orientador(a): Ponczek, Vladimir Pinheiro
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Palavras-chave em Português: School accountability; curriculum-based exams; Exames curriculares
• Link de acesso: http://hdl.handle.net/10438/6552

This dissertation assesses the effects of Brazilian National High School Exam (ENEM) on educational results. This evaluation is based on the theoretical results that curriculum examinations like ENEM are capable of having positive influence on students learning by: raising their effort and pressure on their teachers and school principals. The present work analyses this last channel, through estimating the impact of the public report of ENEM‟s grades by school. The effects of this treatment are estimated on: Math and Reading proficiency, quantities of school inputs and teacher‟s behavior. Two econometric methods are used: i) differences in differences – using data from 2005 (pre treatment period) and 2007 Saeb (first post treatment year) and ii) regression discontinuity design, exploiting the sharp rule of assignment to treatment – only schools with at least 10 students participating of ENEM have their results published. Evidences show that reporting ENEM‟s grades doesn‟t seem to influence the high school students‟ proficiency neither the level of school inputs, result confirmed by both methods. The differences in differences estimates also reveal that teachers of treated schools seem to focus more on solving Math exercises and fixing grammatical concepts, reactions that can be interpreted as adverse effects. This last evidence, however, may be misleading, reflecting only time tendencies and persistent pre treatment differences. We conclude that publication of ENEM‟s grades haven‟t act as an instrument to improve educational results by pressuring teachers and school administrators. Nonetheless, it‟s still possible that this policy works after a longer period or for other types schools, not evaluated by the proposed methods.