• 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.11.2013.tde-13032013-103223
Document
Author
Full name
Marcos André Braz Vaz
E-mail
Institute/School/College
Knowledge Area
Date of Defense
Published
Piracicaba, 2013
Supervisor
Committee
Piedade, Sonia Maria de Stefano (President)
Lara, Idemauro Antonio Rodrigues de
Pião, Antonio Carlos Simões
Title in Portuguese
Estudo de delineamentos experimentais no esquema fatorial duplo com um tratamento adicional
Keywords in Portuguese
Análise de variância
Delineamento experimental
Mínimos quadrados
Regressão linear
Tratamento adicional
Abstract in Portuguese
O presente trabalho teve como objetivo o estudo de experimentos em delineamentos em esquema fatorial duplo com tratamento adicional do tipo testemunha. Para este esquema usa-se a notação A x B +1, em que A representa o primeiro fator com i níveis (i = 1, 2, ..., a) e B representa o segundo fator com j níveis (j = 1, 2, ..., b) com a adição do tratamento adicional. Para a análise de variância deste caso, consideraram-se os modelos lineares yijk = μ + αi + βj + γij + εijk e yh = μ + τ + εh; relacionados, em que yijk é a variável observada no i-ésimo nível do fator α com o j-ésimo nível do fator β da k-ésima repetição (k = 1, 2, ..., r), μ é a média amostral, αi é o efeito do i-ésimo nível do primeiro fator, βj é o efeito do j-ésimo nível do segundo fator, γij é o efeito da interação do i-ésimo nível do fator α com o j-ésimo nível do fator β, εijk é o erro associado independente e identicamente distribuído, εijk~N(0,σ2), yh é a variável observada na h-ésima repetição do tratamento adicional, τ é o efeito do tratamento adicional e εh é o erro associado ao tratamento adicional, independente e identicamente distribuído εh~N(0,σ2). Considerou-se os delineamentos experimentais inteiramente casualizado e blocos casualizados. Para a análise do delineamento em blocos ao acaso, a adição do efeito de blocos λv (v = 1, 2, ..., w) aos modelos, se fez necessária. Foi realizada a dedução da soma de quadrados de tratamentos e sua decomposição para os efeitos dos fatores, sua interação e o contraste com o tratamento adicional. Os graus de liberdade foram deduzidos a partir do posto da matriz núcleo da forma quadrática das somas de quadrados. A técnica do diagrama de Hasse também foi adotada para dedução das somas de quadrados e graus de liberdade. Uma ilustração do método obteve os mesmos resultados da análise de variância do pacote ExpDes no programa R. Curvas de regressão linear foram ajustadas considerando o tratamento controle como um nível de fatores quantitativos. O teste de Dunnett foi empregado para comparar as médias do fatorial com o tratamento controle.
Title in English
Study of experimental design in two-way factorial with an additional treatment
Keywords in English
Additional treatment
Control treatment
Factorial experiment
Interaction
Abstract in English
The present study aimed to study the experiments in two-way factorial designs with additional treatment of type control. For this scheme uses the notation A x B +1, where A represents the first factor levels with i (i = 1, 2, ..., a) and B is the second factor with levels j (j = 1 , 2, ..., b) with the addition of one more treatment. For the analysis of variance of this case, we considered the linear models yijk = μ + αi + βj + γij + εijk and yh = μ + τ + εh; related, wherein yijk is the variable observed in the ith level of factor α with the jth level of factor β of k-th iteration (k = 1, 2, ..., r), μ is the sample mean, αi is the effect of the ith level of the first factor, βj is the effect of the jth level of the second factor, γij is the interaction effect of the ith level of factor α with the jth level of factor β, εijk is the error associated with independent and identically distributed, εijk~N(0,σ2), yh is the variable observed in the hth repetition of the additional treatment, τ is the effect of the additional treatment and εh is the error associated to the additional treatment, independent and identically distributed εh~N(0,σ2). It was considered the completely experimental designs and randomized block design. For the analysis of the randomized block design, the addition of blocks effect λv (v = 1, 2, ..., w) to the models, was necessary. Was performed the deduction of the sum of squares of treatments and their decomposition to the effects of the factors, their interaction and the contrast with the additional treatment. The degrees of freedom were deducted from the posto of the matrix core of the quadratic form of sums of squares. The Hasse diagram technique has also been adopted for deduction of sums of squares and degrees of freedom. An illustration of the method has obtained the same results of analysis of variance program package ExpDes in R. Linear regression analysis was fitted control treatment as a level of the quantitative factors. The Dunnett test was used to compare the means of the factorial with the control treatment.
 
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
2013-03-27
 
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 © 2001-2024. All rights reserved.