A test of separate hypotheses for comparing linear mixed models with non nested fixed effects

Ché L. Smith, Lloyd J. Edwards

Research output: Research - peer-reviewArticle

Abstract

As researchers increasingly rely on linear mixed models to characterize longitudinal data, there is a need for improved techniques for selecting among this class of models which requires specification of both fixed and random effects via a mean model and variance-covariance structure. The process is further complicated when fixed and/or random effects are non nested between models. This paper explores the development of a hypothesis test to compare non nested linear mixed models based on extensions of the work begun by Sir David Cox. We assess the robustness of this approach for comparing models containing correlated measures of body fat for predicting longitudinal cardiometabolic risk.

LanguageEnglish (US)
Pages5487-5500
Number of pages14
JournalCommunications in Statistics - Theory and Methods
Volume46
Issue number11
DOIs
StatePublished - Jun 3 2017

Fingerprint

Test of Hypothesis
Linear Mixed Model
Fixed Effects
Model
Random Effects
Hypothesis Test
Covariance Structure
Longitudinal Data
Model-based
Specification
Robustness
Class

Keywords

  • Bootstrap resampling
  • information criterion
  • linear mixed effects model
  • model selection, non nested

ASJC Scopus subject areas

  • Statistics and Probability

Cite this

A test of separate hypotheses for comparing linear mixed models with non nested fixed effects. / Smith, Ché L.; Edwards, Lloyd J.

In: Communications in Statistics - Theory and Methods, Vol. 46, No. 11, 03.06.2017, p. 5487-5500.

Research output: Research - peer-reviewArticle

@article{884011c2097a4c0b8656367b0c5b462c,
title = "A test of separate hypotheses for comparing linear mixed models with non nested fixed effects",
abstract = "As researchers increasingly rely on linear mixed models to characterize longitudinal data, there is a need for improved techniques for selecting among this class of models which requires specification of both fixed and random effects via a mean model and variance-covariance structure. The process is further complicated when fixed and/or random effects are non nested between models. This paper explores the development of a hypothesis test to compare non nested linear mixed models based on extensions of the work begun by Sir David Cox. We assess the robustness of this approach for comparing models containing correlated measures of body fat for predicting longitudinal cardiometabolic risk.",
keywords = "Bootstrap resampling, information criterion, linear mixed effects model, model selection, non nested",
author = "Smith, {Ché L.} and Edwards, {Lloyd J.}",
year = "2017",
month = "6",
doi = "10.1080/03610926.2015.1104352",
volume = "46",
pages = "5487--5500",
journal = "Communications in Statistics - Theory and Methods",
issn = "0361-0926",
publisher = "Taylor and Francis Ltd.",
number = "11",

}

TY - JOUR

T1 - A test of separate hypotheses for comparing linear mixed models with non nested fixed effects

AU - Smith,Ché L.

AU - Edwards,Lloyd J.

PY - 2017/6/3

Y1 - 2017/6/3

N2 - As researchers increasingly rely on linear mixed models to characterize longitudinal data, there is a need for improved techniques for selecting among this class of models which requires specification of both fixed and random effects via a mean model and variance-covariance structure. The process is further complicated when fixed and/or random effects are non nested between models. This paper explores the development of a hypothesis test to compare non nested linear mixed models based on extensions of the work begun by Sir David Cox. We assess the robustness of this approach for comparing models containing correlated measures of body fat for predicting longitudinal cardiometabolic risk.

AB - As researchers increasingly rely on linear mixed models to characterize longitudinal data, there is a need for improved techniques for selecting among this class of models which requires specification of both fixed and random effects via a mean model and variance-covariance structure. The process is further complicated when fixed and/or random effects are non nested between models. This paper explores the development of a hypothesis test to compare non nested linear mixed models based on extensions of the work begun by Sir David Cox. We assess the robustness of this approach for comparing models containing correlated measures of body fat for predicting longitudinal cardiometabolic risk.

KW - Bootstrap resampling

KW - information criterion

KW - linear mixed effects model

KW - model selection, non nested

UR - http://www.scopus.com/inward/record.url?scp=85013102337&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85013102337&partnerID=8YFLogxK

U2 - 10.1080/03610926.2015.1104352

DO - 10.1080/03610926.2015.1104352

M3 - Article

VL - 46

SP - 5487

EP - 5500

JO - Communications in Statistics - Theory and Methods

T2 - Communications in Statistics - Theory and Methods

JF - Communications in Statistics - Theory and Methods

SN - 0361-0926

IS - 11

ER -