• JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
 
  Bookmark and Share
 
 
Master's Dissertation
DOI
https://doi.org/10.11606/D.45.2009.tde-01032021-140004
Document
Author
Full name
Filipe Jaeger Zabala
E-mail
Institute/School/College
Knowledge Area
Date of Defense
Published
São Paulo, 2009
Supervisor
Committee
Wechsler, Sergio (President)
Migon, Helio dos Santos
Pereira, Carlos Alberto de Braganca
Title in Portuguese
Desempate técnico
Keywords in Portuguese
Eleições
Inferência bayesiana
Pesquisa eleitoral
Proporção
Abstract in Portuguese
Neste trabalho faz-se uma comparação entre as inferências bayesiana e frequentista para a estimação de proporções em um processo eleitoral brasileiro. Foi feito um contraponto entre as probabilidades a posteriori bayesianas e os intervalos/regiões de confiança frequentistas. Mostrou-se que os intervalos de confiança apresentados pelos institutos de pesquisa brasileiros não possuem a confiança que afirmam ter, levando o autor a desenvolver uma metodologia alternativa, denominada neste trabalho como "frequentista correta". De posse das abordagens bayesiana, frequentista correta e dos institutos, encontrou-se um teorema -- em um cenário com dois candidatos -- que relaciona intervalos de confiança frequentistas com probabilidades a posteriori bayesianas. Para três candidatos são apresentados exemplos que contradizem a noção de empate técnico definida pelos institutos, largamente veiculada nos meios de comunicação.
Title in English
Technical tiebreaker
Keywords in English
Bayesian inference
Electoral competition
Electoral survey
Proportion
Abstract in English
In this paper was made comparisons between the bayesian and frequentist inferences for the estimation of the proportions in a Brazilian electoral process. It was made a comparison between the bayesian posterior probabilities and frequentist confidence intervals/regions. It was shown that the confidence intervals submitted by Brazilian research institutes lack the confidence that they claim to have, taking the author to develop an alternative approach, called in this work correct frequentist. With bayesian, correct frequentist and institutes approaches, we derived a theorem that relates in a scenario with two candidates frequentist confidence intervals with bayesian posterior probabilities. For three candidates are given examples that contradict the tie notion defined by the Brazilian research institutes, widely disseminated in mass media.
 
WARNING - Viewing this document is conditioned on your acceptance of the following terms of use:
This document is only for private use for research and teaching activities. Reproduction for commercial use is forbidden. This rights cover the whole data about this document as well as its contents. Any uses or copies of this document in whole or in part must include the author's name.
zabala2009desempate.pdf (609.41 Kbytes)
Publishing Date
2021-03-02
 
WARNING: Learn what derived works are clicking here.
All rights of the thesis/dissertation are from the authors
CeTI-SC/STI
Digital Library of Theses and Dissertations of USP. Copyright © 2001-2023. All rights reserved.