• 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
 
 
Doctoral Thesis
DOI
https://doi.org/10.11606/T.11.2019.tde-20191220-122800
Document
Author
Full name
Francisco das Chagas Estevam da Fonseca
Institute/School/College
Knowledge Area
Date of Defense
Published
Piracicaba, 1999
Supervisor
Title in Portuguese
Medidas repetidas categorizadas incompletas: uma aplicação à confiabilidade de cultivares
Keywords in Portuguese
DADOS CATEGORIZADOS
DELINEAMENTO EXPERIMENTAL
MEDIDAS REPETIDAS
Abstract in Portuguese
Experimentos com medidas repetidas são aqueles nos quais cada indivíduo é observado em diversos tempos diferentes ou sob condições experimentais diferentes. As medidas repetidas são denominadas categorizadas quando a variável resposta é representada por categorias de respostas. Em muitas situações ocorre a existência de dados incompletos. Um dos métodos mais utilizados para a análise de dados incompletos é o de Woolson & Clarke (1984), que assume que os dados são "incompletos por razões totalmente aleatórias" (MCAR - "missing completely at random"), conforme definido em Little & Rubin (1987). Uma aplicação agronômica importante de medidas repetidas categorizadas com dados incompletos foi apresentada por Eskridge (1997). O autor utilizou o conceito de confiabilidade e o método de Woolson & Clarke (1984) para avaliar e comparar híbridos de milho, provenientes de um programa de melhoramento genético. O objetivo do presente trabalho foi aplicar o método de Eskridge (1997), em dois conjuntos de dados, propor um modelo alternativo àquele usado pelo autor, bem como verificar, em cada caso, a suposição de MCAR, utilizando-se o método de Park & Davis ( 1993). Após as análises, concluiu-se que: - em alguns casos o modelo alternativo pode ser mais informativo que o modelo de Eskridge (1997); - o teste do mecanismo de dados incompletos de Park & Davis (1993) é dependente do modelo, não analisando a estrutura, propriamente dita, dos dados.
Title in English
Repeated categorical incomplete measurements: an application to reliability of varieties
Abstract in English
Repeated measures studies are that which each subject is observed in many different times or under different experimental conditions. The repeated measurements are named categorical when the response variable is represented by categories of response. ln much situations occur a existency of missing data. One of the methods more utilized for the analysis of missing data is from the Woolson & Clarke (1984), that assumes the datas are missing completely at random (MCAR), according determined in Little & Rubin (1987). An important agronomic application of repeated categorical measurements with missing data was shown by Eskridge (1997). The author utilized the conceit of reliability and the method of Woolson & Clarke (1984) to evaluate and to compare hybrid of corn, originating of genetic breeding program. The objective of the present work was to apply the method of Eskridge ( 1997), in two sets of data, to propose a alternative model from that used by author, also to verify in each case, the hypothesis of MCAR, utilizing o method of Park & Davis (1993). Thereafter the analysis, was concluded that: - in some cases the alternative model can be more informative than the model of Eskridge (1997); - the test of the missing data mechanism of Park & Davis (1993) is dependent of the model, does not analysing the structure of the data.
 
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.
Publishing Date
2019-12-20
 
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 © 2001-2021. All rights reserved.