Power and Measures of Effect Size in Analysis of Variance with Fixed Versus Random Nested Factors

Matthias Siemer, Jutta Joormann

Research output: Contribution to journalArticle

32 Citations (Scopus)

Abstract

Ignoring a nested factor can influence the validity of statistical decisions about treatment effectiveness. Previous discussions have centered on consequences of ignoring nested factors versus treating them as random factors on Type I errors and measures of effect size (B. E. Wampold & R. C. Serlin, 2000). The authors (a) discuss circumstances under which the treatment of nested provider effects as fixed as opposed to random is appropriate; (b) present 2 formulas for the correct estimation of effect sizes when nested factors are fixed; (c) present the results of Monte Carlo simulations of the consequences of treating providers as fixed versus random on effect size estimates, Type I error rates, and power; and (d) discuss implications of mistaken considerations of provider effects for the study of differential treatment effects in psychotherapy research.

Original languageEnglish
Pages (from-to)497-517
Number of pages21
JournalPsychological Methods
Volume8
Issue number4
DOIs
StatePublished - Dec 1 2003

Fingerprint

Psychotherapy
Analysis of Variance
Research
Power (Psychology)

ASJC Scopus subject areas

  • Psychology(all)

Cite this

Power and Measures of Effect Size in Analysis of Variance with Fixed Versus Random Nested Factors. / Siemer, Matthias; Joormann, Jutta.

In: Psychological Methods, Vol. 8, No. 4, 01.12.2003, p. 497-517.

Research output: Contribution to journalArticle

Siemer, Matthias ; Joormann, Jutta. / Power and Measures of Effect Size in Analysis of Variance with Fixed Versus Random Nested Factors. In: Psychological Methods. 2003 ; Vol. 8, No. 4. pp. 497-517.
@article{1ad29c029841473b8f99285289575b63,
title = "Power and Measures of Effect Size in Analysis of Variance with Fixed Versus Random Nested Factors",
abstract = "Ignoring a nested factor can influence the validity of statistical decisions about treatment effectiveness. Previous discussions have centered on consequences of ignoring nested factors versus treating them as random factors on Type I errors and measures of effect size (B. E. Wampold & R. C. Serlin, 2000). The authors (a) discuss circumstances under which the treatment of nested provider effects as fixed as opposed to random is appropriate; (b) present 2 formulas for the correct estimation of effect sizes when nested factors are fixed; (c) present the results of Monte Carlo simulations of the consequences of treating providers as fixed versus random on effect size estimates, Type I error rates, and power; and (d) discuss implications of mistaken considerations of provider effects for the study of differential treatment effects in psychotherapy research.",
author = "Matthias Siemer and Jutta Joormann",
year = "2003",
month = "12",
day = "1",
doi = "10.1037/1082-989X.8.4.497",
language = "English",
volume = "8",
pages = "497--517",
journal = "Psychological Methods",
issn = "1082-989X",
publisher = "American Psychological Association Inc.",
number = "4",

}

TY - JOUR

T1 - Power and Measures of Effect Size in Analysis of Variance with Fixed Versus Random Nested Factors

AU - Siemer, Matthias

AU - Joormann, Jutta

PY - 2003/12/1

Y1 - 2003/12/1

N2 - Ignoring a nested factor can influence the validity of statistical decisions about treatment effectiveness. Previous discussions have centered on consequences of ignoring nested factors versus treating them as random factors on Type I errors and measures of effect size (B. E. Wampold & R. C. Serlin, 2000). The authors (a) discuss circumstances under which the treatment of nested provider effects as fixed as opposed to random is appropriate; (b) present 2 formulas for the correct estimation of effect sizes when nested factors are fixed; (c) present the results of Monte Carlo simulations of the consequences of treating providers as fixed versus random on effect size estimates, Type I error rates, and power; and (d) discuss implications of mistaken considerations of provider effects for the study of differential treatment effects in psychotherapy research.

AB - Ignoring a nested factor can influence the validity of statistical decisions about treatment effectiveness. Previous discussions have centered on consequences of ignoring nested factors versus treating them as random factors on Type I errors and measures of effect size (B. E. Wampold & R. C. Serlin, 2000). The authors (a) discuss circumstances under which the treatment of nested provider effects as fixed as opposed to random is appropriate; (b) present 2 formulas for the correct estimation of effect sizes when nested factors are fixed; (c) present the results of Monte Carlo simulations of the consequences of treating providers as fixed versus random on effect size estimates, Type I error rates, and power; and (d) discuss implications of mistaken considerations of provider effects for the study of differential treatment effects in psychotherapy research.

UR - http://www.scopus.com/inward/record.url?scp=0346199347&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0346199347&partnerID=8YFLogxK

U2 - 10.1037/1082-989X.8.4.497

DO - 10.1037/1082-989X.8.4.497

M3 - Article

VL - 8

SP - 497

EP - 517

JO - Psychological Methods

JF - Psychological Methods

SN - 1082-989X

IS - 4

ER -