Variable selection in a linear growth curve model with autoregressive within-individual errors

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

In this paper, we consider variable selection schemes in a linear random coefficient growth curve model. Unbalanced within-individual data is assumed to have first-order linear correlation (ρ). A variable selection scheme suggested in the between-individual model is a weighted stepwise selection. A simulation study is conducted to compare the performances of the maximum likelihood (ML) and the ordinary least square (OLS) estimators in the within-individual regression model in terms of the ability of variable selection in the between-individual model. Simulation results indicate the following: Under the suspicion of high autocorrelation error (ρ > 0.5), the ML estimation is necessary in the within-individual model. When it is believed that the p is as small as 0.1 and the heterogeneity of variance is relatively low then, the OLS estimator can replace the ML estimator in terms of selection performance in the between-individual model.

Original languageEnglish
Pages (from-to)247-255
Number of pages9
JournalJournal of Statistical Computation and Simulation
Volume40
Issue number3-4
DOIs
Publication statusPublished - 1992 Apr 1

Fingerprint

Growth Curve Model
Variable Selection
Ordinary Least Squares Estimator
Selection of Variables
Maximum likelihood
Random Coefficients
Autocorrelation
Maximum Likelihood Estimation
Model
Maximum Likelihood Estimator
Maximum Likelihood
Regression Model
Maximum likelihood estimation
Simulation Study
First-order
Necessary
Individual model
Growth curve
Variable selection
Simulation

All Science Journal Classification (ASJC) codes

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

Cite this

@article{48d2f640e155464486630c83795362b4,
title = "Variable selection in a linear growth curve model with autoregressive within-individual errors",
abstract = "In this paper, we consider variable selection schemes in a linear random coefficient growth curve model. Unbalanced within-individual data is assumed to have first-order linear correlation (ρ). A variable selection scheme suggested in the between-individual model is a weighted stepwise selection. A simulation study is conducted to compare the performances of the maximum likelihood (ML) and the ordinary least square (OLS) estimators in the within-individual regression model in terms of the ability of variable selection in the between-individual model. Simulation results indicate the following: Under the suspicion of high autocorrelation error (ρ > 0.5), the ML estimation is necessary in the within-individual model. When it is believed that the p is as small as 0.1 and the heterogeneity of variance is relatively low then, the OLS estimator can replace the ML estimator in terms of selection performance in the between-individual model.",
author = "Sohn, {So Young}",
year = "1992",
month = "4",
day = "1",
doi = "10.1080/00949659208811380",
language = "English",
volume = "40",
pages = "247--255",
journal = "Journal of Statistical Computation and Simulation",
issn = "0094-9655",
publisher = "Taylor and Francis Ltd.",
number = "3-4",

}

TY - JOUR

T1 - Variable selection in a linear growth curve model with autoregressive within-individual errors

AU - Sohn, So Young

PY - 1992/4/1

Y1 - 1992/4/1

N2 - In this paper, we consider variable selection schemes in a linear random coefficient growth curve model. Unbalanced within-individual data is assumed to have first-order linear correlation (ρ). A variable selection scheme suggested in the between-individual model is a weighted stepwise selection. A simulation study is conducted to compare the performances of the maximum likelihood (ML) and the ordinary least square (OLS) estimators in the within-individual regression model in terms of the ability of variable selection in the between-individual model. Simulation results indicate the following: Under the suspicion of high autocorrelation error (ρ > 0.5), the ML estimation is necessary in the within-individual model. When it is believed that the p is as small as 0.1 and the heterogeneity of variance is relatively low then, the OLS estimator can replace the ML estimator in terms of selection performance in the between-individual model.

AB - In this paper, we consider variable selection schemes in a linear random coefficient growth curve model. Unbalanced within-individual data is assumed to have first-order linear correlation (ρ). A variable selection scheme suggested in the between-individual model is a weighted stepwise selection. A simulation study is conducted to compare the performances of the maximum likelihood (ML) and the ordinary least square (OLS) estimators in the within-individual regression model in terms of the ability of variable selection in the between-individual model. Simulation results indicate the following: Under the suspicion of high autocorrelation error (ρ > 0.5), the ML estimation is necessary in the within-individual model. When it is believed that the p is as small as 0.1 and the heterogeneity of variance is relatively low then, the OLS estimator can replace the ML estimator in terms of selection performance in the between-individual model.

UR - http://www.scopus.com/inward/record.url?scp=0344762052&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0344762052&partnerID=8YFLogxK

U2 - 10.1080/00949659208811380

DO - 10.1080/00949659208811380

M3 - Article

AN - SCOPUS:0344762052

VL - 40

SP - 247

EP - 255

JO - Journal of Statistical Computation and Simulation

JF - Journal of Statistical Computation and Simulation

SN - 0094-9655

IS - 3-4

ER -