Local influence to detect influential data structures for generalized linear mixed models.

M.J.N. Ouwens, E.S. Tan, M.P.F. Berger

Research output: Contribution to journalArticleAcademicpeer-review

24 Citations (Scopus)

Abstract

Biometrics 2001 Dec;57(4):1166-72 Related Articles, Books, LinkOut


Local influence to detect influential data structures for generalized linear mixed models.

Ouwens MJ, Tan FE, Berger MP.

Department of Methodology and Statistics, Maastricht University, The Netherlands. mario.ouwens@stat.unimaas.nl

This article discusses the generalization of the local influence measures for normally distributed responses to local influence measures for generalized linear models with random effects. For these models, it is shown that the subject-oriented influence measure is a special case of the proposed observation-oriented influence measure. A two-step diagnostic procedure is proposed. The first step is to search for influential subjects. A search for influential observations is proposed as the second step. An illustration of a two-treatment, multiple-period crossover trial demonstrates the practical importance of the detection of influential observations in addition to the detection of influential subjects.

Original languageEnglish
Pages (from-to)1166-1172
Number of pages7
JournalBiometrics
Volume57
Issue number4
DOIs
Publication statusPublished - 1 Jan 2001

Cite this

Ouwens, M.J.N. ; Tan, E.S. ; Berger, M.P.F. / Local influence to detect influential data structures for generalized linear mixed models. In: Biometrics. 2001 ; Vol. 57, No. 4. pp. 1166-1172.
@article{ed625ac07ada4ab7a13c8cc30249c97e,
title = "Local influence to detect influential data structures for generalized linear mixed models.",
abstract = "Biometrics 2001 Dec;57(4):1166-72 Related Articles, Books, LinkOut Local influence to detect influential data structures for generalized linear mixed models.Ouwens MJ, Tan FE, Berger MP.Department of Methodology and Statistics, Maastricht University, The Netherlands. mario.ouwens@stat.unimaas.nlThis article discusses the generalization of the local influence measures for normally distributed responses to local influence measures for generalized linear models with random effects. For these models, it is shown that the subject-oriented influence measure is a special case of the proposed observation-oriented influence measure. A two-step diagnostic procedure is proposed. The first step is to search for influential subjects. A search for influential observations is proposed as the second step. An illustration of a two-treatment, multiple-period crossover trial demonstrates the practical importance of the detection of influential observations in addition to the detection of influential subjects.",
author = "M.J.N. Ouwens and E.S. Tan and M.P.F. Berger",
year = "2001",
month = "1",
day = "1",
doi = "10.1111/j.0006-341X.2001.01166.x",
language = "English",
volume = "57",
pages = "1166--1172",
journal = "Biometrics",
issn = "0006-341X",
publisher = "Wiley-Blackwell",
number = "4",

}

Local influence to detect influential data structures for generalized linear mixed models. / Ouwens, M.J.N.; Tan, E.S.; Berger, M.P.F.

In: Biometrics, Vol. 57, No. 4, 01.01.2001, p. 1166-1172.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Local influence to detect influential data structures for generalized linear mixed models.

AU - Ouwens, M.J.N.

AU - Tan, E.S.

AU - Berger, M.P.F.

PY - 2001/1/1

Y1 - 2001/1/1

N2 - Biometrics 2001 Dec;57(4):1166-72 Related Articles, Books, LinkOut Local influence to detect influential data structures for generalized linear mixed models.Ouwens MJ, Tan FE, Berger MP.Department of Methodology and Statistics, Maastricht University, The Netherlands. mario.ouwens@stat.unimaas.nlThis article discusses the generalization of the local influence measures for normally distributed responses to local influence measures for generalized linear models with random effects. For these models, it is shown that the subject-oriented influence measure is a special case of the proposed observation-oriented influence measure. A two-step diagnostic procedure is proposed. The first step is to search for influential subjects. A search for influential observations is proposed as the second step. An illustration of a two-treatment, multiple-period crossover trial demonstrates the practical importance of the detection of influential observations in addition to the detection of influential subjects.

AB - Biometrics 2001 Dec;57(4):1166-72 Related Articles, Books, LinkOut Local influence to detect influential data structures for generalized linear mixed models.Ouwens MJ, Tan FE, Berger MP.Department of Methodology and Statistics, Maastricht University, The Netherlands. mario.ouwens@stat.unimaas.nlThis article discusses the generalization of the local influence measures for normally distributed responses to local influence measures for generalized linear models with random effects. For these models, it is shown that the subject-oriented influence measure is a special case of the proposed observation-oriented influence measure. A two-step diagnostic procedure is proposed. The first step is to search for influential subjects. A search for influential observations is proposed as the second step. An illustration of a two-treatment, multiple-period crossover trial demonstrates the practical importance of the detection of influential observations in addition to the detection of influential subjects.

U2 - 10.1111/j.0006-341X.2001.01166.x

DO - 10.1111/j.0006-341X.2001.01166.x

M3 - Article

VL - 57

SP - 1166

EP - 1172

JO - Biometrics

JF - Biometrics

SN - 0006-341X

IS - 4

ER -