Power analyses for longitudinal trials and other clustered designs

Xin M. Tu, J. Kowalski, J. Zhang, K. G. Lynch, P. Crits-Christoph

Research output: Contribution to journalArticle

Abstract

Existing methods for power and sample size estimation for longitudinal and other clustered study designs have limited applications. In this paper, we review and extend existing approaches to improve these limitations. In particular, we focus on power analysis for the two most popular approaches for clustered data analysis, the generalized estimating equations and the linear mixed-effects models. By basing the derivation of the power function on the asymptotic distribution of the model estimates, the proposed approach provides estimates of power that are consistent with the methods of inference for data analysis. The proposed methodology is illustrated with numerous examples that are motivated by real study designs.

Original languageEnglish (US)
Pages (from-to)2799-2815
Number of pages17
JournalStatistics in Medicine
Volume23
Issue number18
DOIs
StatePublished - Sep 30 2004

Fingerprint

Data analysis
Linear Mixed Effects Model
Clustered Data
Generalized Estimating Equations
Power Analysis
Power Function
Sample Size
Estimate
Asymptotic distribution
Methodology
Design
Model
Review

Keywords

  • Epidemiological study
  • GEE
  • HIV
  • Intraclass correlation
  • Linear mixed-effects models
  • Psychosocial and survey research

ASJC Scopus subject areas

  • Epidemiology

Cite this

Tu, X. M., Kowalski, J., Zhang, J., Lynch, K. G., & Crits-Christoph, P. (2004). Power analyses for longitudinal trials and other clustered designs. Statistics in Medicine, 23(18), 2799-2815. https://doi.org/10.1002/sim.1869

Power analyses for longitudinal trials and other clustered designs. / Tu, Xin M.; Kowalski, J.; Zhang, J.; Lynch, K. G.; Crits-Christoph, P.

In: Statistics in Medicine, Vol. 23, No. 18, 30.09.2004, p. 2799-2815.

Research output: Contribution to journalArticle

Tu, XM, Kowalski, J, Zhang, J, Lynch, KG & Crits-Christoph, P 2004, 'Power analyses for longitudinal trials and other clustered designs', Statistics in Medicine, vol. 23, no. 18, pp. 2799-2815. https://doi.org/10.1002/sim.1869
Tu XM, Kowalski J, Zhang J, Lynch KG, Crits-Christoph P. Power analyses for longitudinal trials and other clustered designs. Statistics in Medicine. 2004 Sep 30;23(18):2799-2815. https://doi.org/10.1002/sim.1869
Tu, Xin M. ; Kowalski, J. ; Zhang, J. ; Lynch, K. G. ; Crits-Christoph, P. / Power analyses for longitudinal trials and other clustered designs. In: Statistics in Medicine. 2004 ; Vol. 23, No. 18. pp. 2799-2815.
@article{628c974f8ffd49ccaaccdd6e14b167f1,
title = "Power analyses for longitudinal trials and other clustered designs",
abstract = "Existing methods for power and sample size estimation for longitudinal and other clustered study designs have limited applications. In this paper, we review and extend existing approaches to improve these limitations. In particular, we focus on power analysis for the two most popular approaches for clustered data analysis, the generalized estimating equations and the linear mixed-effects models. By basing the derivation of the power function on the asymptotic distribution of the model estimates, the proposed approach provides estimates of power that are consistent with the methods of inference for data analysis. The proposed methodology is illustrated with numerous examples that are motivated by real study designs.",
keywords = "Epidemiological study, GEE, HIV, Intraclass correlation, Linear mixed-effects models, Psychosocial and survey research",
author = "Tu, {Xin M.} and J. Kowalski and J. Zhang and Lynch, {K. G.} and P. Crits-Christoph",
year = "2004",
month = "9",
day = "30",
doi = "10.1002/sim.1869",
language = "English (US)",
volume = "23",
pages = "2799--2815",
journal = "Statistics in Medicine",
issn = "0277-6715",
publisher = "John Wiley and Sons Ltd",
number = "18",

}

TY - JOUR

T1 - Power analyses for longitudinal trials and other clustered designs

AU - Tu, Xin M.

AU - Kowalski, J.

AU - Zhang, J.

AU - Lynch, K. G.

AU - Crits-Christoph, P.

PY - 2004/9/30

Y1 - 2004/9/30

N2 - Existing methods for power and sample size estimation for longitudinal and other clustered study designs have limited applications. In this paper, we review and extend existing approaches to improve these limitations. In particular, we focus on power analysis for the two most popular approaches for clustered data analysis, the generalized estimating equations and the linear mixed-effects models. By basing the derivation of the power function on the asymptotic distribution of the model estimates, the proposed approach provides estimates of power that are consistent with the methods of inference for data analysis. The proposed methodology is illustrated with numerous examples that are motivated by real study designs.

AB - Existing methods for power and sample size estimation for longitudinal and other clustered study designs have limited applications. In this paper, we review and extend existing approaches to improve these limitations. In particular, we focus on power analysis for the two most popular approaches for clustered data analysis, the generalized estimating equations and the linear mixed-effects models. By basing the derivation of the power function on the asymptotic distribution of the model estimates, the proposed approach provides estimates of power that are consistent with the methods of inference for data analysis. The proposed methodology is illustrated with numerous examples that are motivated by real study designs.

KW - Epidemiological study

KW - GEE

KW - HIV

KW - Intraclass correlation

KW - Linear mixed-effects models

KW - Psychosocial and survey research

UR - http://www.scopus.com/inward/record.url?scp=4444219777&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=4444219777&partnerID=8YFLogxK

U2 - 10.1002/sim.1869

DO - 10.1002/sim.1869

M3 - Article

C2 - 15344187

AN - SCOPUS:4444219777

VL - 23

SP - 2799

EP - 2815

JO - Statistics in Medicine

JF - Statistics in Medicine

SN - 0277-6715

IS - 18

ER -