<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"> <html> <head> <title></title> </head> <body style="font-family:Arial;font-size:14px"> Ciclo de Palestras do Programa de Pós-graduação em Estatística da UnB Divulgação de palestra do Programa de Pós-graduação em Estatística da UnB. Palestrante: Alexandra M. Schmidt (IM-UFRJ) DATA: 10/09/2015 (quinta-feira) HORÁRIO: 10:30h LOCAL: Sala Multiuso EST-UnB Titulo: A Hierarchical Dynamic Beta Regression Model of School Performance in the Brazilian Mathematical Olympiads for Public Schools Resumo: The Brazilian Mathematical Olympiads for Public Schools (Olimpíada Brasileira de Matemática em Escolas Públicas or OBMEP) is held every year since 2005. In the 2013 edition there were over 47,000 schools registered involving nearly 19.2 million students. The Brazilian public educational system is divided into three different administrative levels: federal, state and municipal. In OBMEP students are divided into three different educational levels. We aim at studying the performance of schools, nationwide, which have been taking part of OBMEP from 2006 until 2013. Understanding what covariates influence the performance of schools in the different levels of OBMEP is important as it might help defining, or revising, strategies about teaching mathematics, attracting more students to the area, besides being useful to guide public policy of resource allocation. We propose a hierarchical dynamic beta regression model that captures effects across different levels and years. We also allow the precision parameter of the beta distribution to be a function of covariates whose effects evolve smoothly with time according to a hierarchical dynamic model. Results show that schools with greater proportion of boys tend to have better performance in the second and third educational levels. However, gender does not seem to affect the performance of schools in the first level. Regardless of the educational level, federal schools show higher performance than municipal or state schools. The human development index (HDI) of the municipality the school belongs to affects positively the performance of schools independent of the educational level considered. This is joint work with Caroline P. de Moraes, and Helio S. Migon. </body> </html>