Quantile dispersion graphs for the comparison of designs for a random two-way model

Juneyoung Lee, André I. Khuri

Research output: Contribution to journalArticle

8 Citations (Scopus)

Abstract

The choice of optimal designs for the estimation of variance components depends on the method of estimation, the model used, and the values of the variance components themselves. Traditional comparisons of such designs cannot therefore be made without some prior knowledge of the variance components. This problem was circumvented by the introduction of the so-called quantile dispersion graphs (QDGs) in Khuri (1997) and Lee and Khuri (1999). The present article provides an extension of the use of the QDGs to the comparison of designs for an unbalanced random two-way model without interaction. Two methods of estimation of the variance components are considered, namely, the analysis of variance (ANOVA) and maximum likelihood (ML).

Original languageEnglish
Pages (from-to)123-137
Number of pages15
JournalJournal of Statistical Planning and Inference
Volume91
Issue number1
Publication statusPublished - 2000 Nov 1
Externally publishedYes

Fingerprint

Variance Components
Quantile
Graph in graph theory
Components of Variance
Analysis of variance
Analysis of variance (ANOVA)
Prior Knowledge
Maximum likelihood
Maximum Likelihood
Model
Interaction
Design
Graph
Variance components

Keywords

  • ANOVA estimation
  • Davies' algorithm
  • Maximum likelihood estimation
  • Primary 62J10
  • Secondary 62K99
  • Unbalanced design
  • Variance components

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty
  • Applied Mathematics

Cite this

Quantile dispersion graphs for the comparison of designs for a random two-way model. / Lee, Juneyoung; Khuri, André I.

In: Journal of Statistical Planning and Inference, Vol. 91, No. 1, 01.11.2000, p. 123-137.

Research output: Contribution to journalArticle

@article{9cfcb090a5884543b2a32119b7428963,
title = "Quantile dispersion graphs for the comparison of designs for a random two-way model",
abstract = "The choice of optimal designs for the estimation of variance components depends on the method of estimation, the model used, and the values of the variance components themselves. Traditional comparisons of such designs cannot therefore be made without some prior knowledge of the variance components. This problem was circumvented by the introduction of the so-called quantile dispersion graphs (QDGs) in Khuri (1997) and Lee and Khuri (1999). The present article provides an extension of the use of the QDGs to the comparison of designs for an unbalanced random two-way model without interaction. Two methods of estimation of the variance components are considered, namely, the analysis of variance (ANOVA) and maximum likelihood (ML).",
keywords = "ANOVA estimation, Davies' algorithm, Maximum likelihood estimation, Primary 62J10, Secondary 62K99, Unbalanced design, Variance components",
author = "Juneyoung Lee and Khuri, {Andr{\'e} I.}",
year = "2000",
month = "11",
day = "1",
language = "English",
volume = "91",
pages = "123--137",
journal = "Journal of Statistical Planning and Inference",
issn = "0378-3758",
publisher = "Elsevier",
number = "1",

}

TY - JOUR

T1 - Quantile dispersion graphs for the comparison of designs for a random two-way model

AU - Lee, Juneyoung

AU - Khuri, André I.

PY - 2000/11/1

Y1 - 2000/11/1

N2 - The choice of optimal designs for the estimation of variance components depends on the method of estimation, the model used, and the values of the variance components themselves. Traditional comparisons of such designs cannot therefore be made without some prior knowledge of the variance components. This problem was circumvented by the introduction of the so-called quantile dispersion graphs (QDGs) in Khuri (1997) and Lee and Khuri (1999). The present article provides an extension of the use of the QDGs to the comparison of designs for an unbalanced random two-way model without interaction. Two methods of estimation of the variance components are considered, namely, the analysis of variance (ANOVA) and maximum likelihood (ML).

AB - The choice of optimal designs for the estimation of variance components depends on the method of estimation, the model used, and the values of the variance components themselves. Traditional comparisons of such designs cannot therefore be made without some prior knowledge of the variance components. This problem was circumvented by the introduction of the so-called quantile dispersion graphs (QDGs) in Khuri (1997) and Lee and Khuri (1999). The present article provides an extension of the use of the QDGs to the comparison of designs for an unbalanced random two-way model without interaction. Two methods of estimation of the variance components are considered, namely, the analysis of variance (ANOVA) and maximum likelihood (ML).

KW - ANOVA estimation

KW - Davies' algorithm

KW - Maximum likelihood estimation

KW - Primary 62J10

KW - Secondary 62K99

KW - Unbalanced design

KW - Variance components

UR - http://www.scopus.com/inward/record.url?scp=0001949387&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0001949387&partnerID=8YFLogxK

M3 - Article

VL - 91

SP - 123

EP - 137

JO - Journal of Statistical Planning and Inference

JF - Journal of Statistical Planning and Inference

SN - 0378-3758

IS - 1

ER -