Sample sizes required to detect two-way and three-way interactions involving slope differences in mixed-effects linear models

Moonseong Heo, Andrew C. Leon

Research output: Contribution to journalArticle

27 Citations (Scopus)

Abstract

Based on maximum likelihood estimates obtained from mixed-effects linear models, closed-form power functions are derived to detect two-way and three-way interactions that involve longitudinal course of outcome over time in clinical trials. Sample size estimates are shown to decrease with increasing within-subject correlations. It is further shown that when clinical trial designs are balanced in group sizes, the sample size required to detect an effect size for a three-way interaction is exactly fourfold that required to detect the same effect size of a two-way interaction. Furthermore, this fourfold relationship virtually holds for unbalanced allocations of subjects if one factor is balanced in the three-way interaction model. Simulations are presented that verify the sample size estimates for two-way and three-way interactions.

Original languageEnglish (US)
Pages (from-to)787-802
Number of pages16
JournalJournal of Biopharmaceutical Statistics
Volume20
Issue number4
DOIs
StatePublished - Jul 2010

Fingerprint

Linear Mixed Effects Model
Sample Size
Linear Models
Slope
Interaction
Effect Size
Clinical Trials
Likelihood Functions
Power Function
Maximum Likelihood Estimate
Estimate
Closed-form
Verify
Decrease
Simulation

Keywords

  • Clinical trials
  • Effect size
  • Power function
  • Sample size requirements
  • Three-way interaction
  • Two-way interaction

ASJC Scopus subject areas

  • Pharmacology (medical)
  • Pharmacology
  • Statistics and Probability

Cite this

Sample sizes required to detect two-way and three-way interactions involving slope differences in mixed-effects linear models. / Heo, Moonseong; Leon, Andrew C.

In: Journal of Biopharmaceutical Statistics, Vol. 20, No. 4, 07.2010, p. 787-802.

Research output: Contribution to journalArticle

@article{4829a63e8aec44e7a8c441707feee6b0,
title = "Sample sizes required to detect two-way and three-way interactions involving slope differences in mixed-effects linear models",
abstract = "Based on maximum likelihood estimates obtained from mixed-effects linear models, closed-form power functions are derived to detect two-way and three-way interactions that involve longitudinal course of outcome over time in clinical trials. Sample size estimates are shown to decrease with increasing within-subject correlations. It is further shown that when clinical trial designs are balanced in group sizes, the sample size required to detect an effect size for a three-way interaction is exactly fourfold that required to detect the same effect size of a two-way interaction. Furthermore, this fourfold relationship virtually holds for unbalanced allocations of subjects if one factor is balanced in the three-way interaction model. Simulations are presented that verify the sample size estimates for two-way and three-way interactions.",
keywords = "Clinical trials, Effect size, Power function, Sample size requirements, Three-way interaction, Two-way interaction",
author = "Moonseong Heo and Leon, {Andrew C.}",
year = "2010",
month = "7",
doi = "10.1080/10543401003618819",
language = "English (US)",
volume = "20",
pages = "787--802",
journal = "Journal of Biopharmaceutical Statistics",
issn = "1054-3406",
publisher = "Taylor and Francis Ltd.",
number = "4",

}

TY - JOUR

T1 - Sample sizes required to detect two-way and three-way interactions involving slope differences in mixed-effects linear models

AU - Heo, Moonseong

AU - Leon, Andrew C.

PY - 2010/7

Y1 - 2010/7

N2 - Based on maximum likelihood estimates obtained from mixed-effects linear models, closed-form power functions are derived to detect two-way and three-way interactions that involve longitudinal course of outcome over time in clinical trials. Sample size estimates are shown to decrease with increasing within-subject correlations. It is further shown that when clinical trial designs are balanced in group sizes, the sample size required to detect an effect size for a three-way interaction is exactly fourfold that required to detect the same effect size of a two-way interaction. Furthermore, this fourfold relationship virtually holds for unbalanced allocations of subjects if one factor is balanced in the three-way interaction model. Simulations are presented that verify the sample size estimates for two-way and three-way interactions.

AB - Based on maximum likelihood estimates obtained from mixed-effects linear models, closed-form power functions are derived to detect two-way and three-way interactions that involve longitudinal course of outcome over time in clinical trials. Sample size estimates are shown to decrease with increasing within-subject correlations. It is further shown that when clinical trial designs are balanced in group sizes, the sample size required to detect an effect size for a three-way interaction is exactly fourfold that required to detect the same effect size of a two-way interaction. Furthermore, this fourfold relationship virtually holds for unbalanced allocations of subjects if one factor is balanced in the three-way interaction model. Simulations are presented that verify the sample size estimates for two-way and three-way interactions.

KW - Clinical trials

KW - Effect size

KW - Power function

KW - Sample size requirements

KW - Three-way interaction

KW - Two-way interaction

UR - http://www.scopus.com/inward/record.url?scp=77952904258&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77952904258&partnerID=8YFLogxK

U2 - 10.1080/10543401003618819

DO - 10.1080/10543401003618819

M3 - Article

C2 - 20496206

AN - SCOPUS:77952904258

VL - 20

SP - 787

EP - 802

JO - Journal of Biopharmaceutical Statistics

JF - Journal of Biopharmaceutical Statistics

SN - 1054-3406

IS - 4

ER -