Model selection in linear mixed effect models

Heng Peng, Ying Lu

Research output: Contribution to journalArticle

Abstract

Mixed effect models are fundamental tools for the analysis of longitudinal data, panel data and cross-sectional data. They are widely used by various fields of social sciences, medical and biological sciences. However, the complex nature of these models has made variable selection and parameter estimation a challenging problem. In this paper, we propose a simple iterative procedure that estimates and selects fixed and random effects for linear mixed models. In particular, we propose to utilize the partial consistency property of the random effect coefficients and select groups of random effects simultaneously via a data-oriented penalty function (the smoothly clipped absolute deviation penalty function). We show that the proposed method is a consistent variable selection procedure and possesses some oracle properties. Simulation studies and a real data analysis are also conducted to empirically examine the performance of this procedure.

Original languageEnglish (US)
Pages (from-to)109-129
Number of pages21
JournalJournal of Multivariate Analysis
Volume109
DOIs
StatePublished - Aug 2012

Fingerprint

Linear Mixed Effects Model
Random Effects
Model Selection
Penalty Function
Variable Selection
Oracle Property
Mixed Effects Model
Linear Mixed Model
Social sciences
Fixed Effects
Panel Data
Selection Procedures
Social Sciences
Longitudinal Data
Iterative Procedure
Parameter estimation
Parameter Estimation
Data analysis
Deviation
Simulation Study

Keywords

  • Group selection
  • Model selection
  • Oracle property
  • Penalized least squares
  • SCAD function

ASJC Scopus subject areas

  • Statistics, Probability and Uncertainty
  • Numerical Analysis
  • Statistics and Probability

Cite this

Model selection in linear mixed effect models. / Peng, Heng; Lu, Ying.

In: Journal of Multivariate Analysis, Vol. 109, 08.2012, p. 109-129.

Research output: Contribution to journalArticle

@article{cfdb5aecb37e44dba1f3c28195af0c6e,
title = "Model selection in linear mixed effect models",
abstract = "Mixed effect models are fundamental tools for the analysis of longitudinal data, panel data and cross-sectional data. They are widely used by various fields of social sciences, medical and biological sciences. However, the complex nature of these models has made variable selection and parameter estimation a challenging problem. In this paper, we propose a simple iterative procedure that estimates and selects fixed and random effects for linear mixed models. In particular, we propose to utilize the partial consistency property of the random effect coefficients and select groups of random effects simultaneously via a data-oriented penalty function (the smoothly clipped absolute deviation penalty function). We show that the proposed method is a consistent variable selection procedure and possesses some oracle properties. Simulation studies and a real data analysis are also conducted to empirically examine the performance of this procedure.",
keywords = "Group selection, Model selection, Oracle property, Penalized least squares, SCAD function",
author = "Heng Peng and Ying Lu",
year = "2012",
month = "8",
doi = "10.1016/j.jmva.2012.02.005",
language = "English (US)",
volume = "109",
pages = "109--129",
journal = "Journal of Multivariate Analysis",
issn = "0047-259X",
publisher = "Academic Press Inc.",

}

TY - JOUR

T1 - Model selection in linear mixed effect models

AU - Peng, Heng

AU - Lu, Ying

PY - 2012/8

Y1 - 2012/8

N2 - Mixed effect models are fundamental tools for the analysis of longitudinal data, panel data and cross-sectional data. They are widely used by various fields of social sciences, medical and biological sciences. However, the complex nature of these models has made variable selection and parameter estimation a challenging problem. In this paper, we propose a simple iterative procedure that estimates and selects fixed and random effects for linear mixed models. In particular, we propose to utilize the partial consistency property of the random effect coefficients and select groups of random effects simultaneously via a data-oriented penalty function (the smoothly clipped absolute deviation penalty function). We show that the proposed method is a consistent variable selection procedure and possesses some oracle properties. Simulation studies and a real data analysis are also conducted to empirically examine the performance of this procedure.

AB - Mixed effect models are fundamental tools for the analysis of longitudinal data, panel data and cross-sectional data. They are widely used by various fields of social sciences, medical and biological sciences. However, the complex nature of these models has made variable selection and parameter estimation a challenging problem. In this paper, we propose a simple iterative procedure that estimates and selects fixed and random effects for linear mixed models. In particular, we propose to utilize the partial consistency property of the random effect coefficients and select groups of random effects simultaneously via a data-oriented penalty function (the smoothly clipped absolute deviation penalty function). We show that the proposed method is a consistent variable selection procedure and possesses some oracle properties. Simulation studies and a real data analysis are also conducted to empirically examine the performance of this procedure.

KW - Group selection

KW - Model selection

KW - Oracle property

KW - Penalized least squares

KW - SCAD function

UR - http://www.scopus.com/inward/record.url?scp=84860437900&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84860437900&partnerID=8YFLogxK

U2 - 10.1016/j.jmva.2012.02.005

DO - 10.1016/j.jmva.2012.02.005

M3 - Article

VL - 109

SP - 109

EP - 129

JO - Journal of Multivariate Analysis

JF - Journal of Multivariate Analysis

SN - 0047-259X

ER -