Adjustment of the profile likelihood for a class of normal regression models

Maria Durban, Iain D. Currie

Research output: Contribution to journalArticle

Abstract

We apply the method of McCullagh & Tibshirani (1990) to a generalization of the model for variance components in which the parameter of interest can appear in both the mean and variance. We obtain the exact adjusted profile log-likelihood score function. For the variance components model, we obtain the adjusted profile log-likelihood and show that it equals the restricted log-likelihood of Patterson & Thompson (1971). We discuss an example due to Kempton (1982) of a regression model with autoregressive terms in which the parameter of interest appears in both the mean and variance.

Original languageEnglish
Pages (from-to)535-540
Number of pages6
JournalScandinavian Journal of Statistics
Volume27
Issue number3
Publication statusPublished - Sep 2000

Fingerprint

Profile Likelihood
Likelihood
Regression Model
Adjustment
Variance Component Model
Score Function
Variance Components
Likelihood Function
Term
Profile
Class
Model
Generalization

Keywords

  • Competition
  • Mixed model
  • Profile likelihood
  • REML
  • Score function
  • Variance components

Cite this

Durban, Maria ; Currie, Iain D. / Adjustment of the profile likelihood for a class of normal regression models. In: Scandinavian Journal of Statistics. 2000 ; Vol. 27, No. 3. pp. 535-540.
@article{add1752042cb466e951448bd6f093b7f,
title = "Adjustment of the profile likelihood for a class of normal regression models",
abstract = "We apply the method of McCullagh & Tibshirani (1990) to a generalization of the model for variance components in which the parameter of interest can appear in both the mean and variance. We obtain the exact adjusted profile log-likelihood score function. For the variance components model, we obtain the adjusted profile log-likelihood and show that it equals the restricted log-likelihood of Patterson & Thompson (1971). We discuss an example due to Kempton (1982) of a regression model with autoregressive terms in which the parameter of interest appears in both the mean and variance.",
keywords = "Competition, Mixed model, Profile likelihood, REML, Score function, Variance components",
author = "Maria Durban and Currie, {Iain D.}",
year = "2000",
month = "9",
language = "English",
volume = "27",
pages = "535--540",
journal = "Scandinavian Journal of Statistics",
issn = "0303-6898",
publisher = "Wiley-Blackwell",
number = "3",

}

Adjustment of the profile likelihood for a class of normal regression models. / Durban, Maria; Currie, Iain D.

In: Scandinavian Journal of Statistics, Vol. 27, No. 3, 09.2000, p. 535-540.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Adjustment of the profile likelihood for a class of normal regression models

AU - Durban, Maria

AU - Currie, Iain D.

PY - 2000/9

Y1 - 2000/9

N2 - We apply the method of McCullagh & Tibshirani (1990) to a generalization of the model for variance components in which the parameter of interest can appear in both the mean and variance. We obtain the exact adjusted profile log-likelihood score function. For the variance components model, we obtain the adjusted profile log-likelihood and show that it equals the restricted log-likelihood of Patterson & Thompson (1971). We discuss an example due to Kempton (1982) of a regression model with autoregressive terms in which the parameter of interest appears in both the mean and variance.

AB - We apply the method of McCullagh & Tibshirani (1990) to a generalization of the model for variance components in which the parameter of interest can appear in both the mean and variance. We obtain the exact adjusted profile log-likelihood score function. For the variance components model, we obtain the adjusted profile log-likelihood and show that it equals the restricted log-likelihood of Patterson & Thompson (1971). We discuss an example due to Kempton (1982) of a regression model with autoregressive terms in which the parameter of interest appears in both the mean and variance.

KW - Competition

KW - Mixed model

KW - Profile likelihood

KW - REML

KW - Score function

KW - Variance components

UR - http://www.scopus.com/inward/record.url?scp=0034336106&partnerID=8YFLogxK

M3 - Article

VL - 27

SP - 535

EP - 540

JO - Scandinavian Journal of Statistics

JF - Scandinavian Journal of Statistics

SN - 0303-6898

IS - 3

ER -