Evaluating the robustness of repeated measures analyses : the case of small sample sizes and nonnormal data

Publikationstyp:
Zeitschriftenaufsatz
Metadaten:
Autoren
  • Daniel Oberfeld-Twistel
  • Thomas Franke
Sammlungen
  • metadata
ISSN
1554-3528
Ausgabe der Veröffentlichung
3
Zeitschrift
Behavior research methods
Schlüsselwörter
  • 150 Psychologie
  • 150 Psychology
Sprache
eng
Paginierung
Seiten: 792 - 812
Datum der Veröffentlichung
2013
Herausgeber
Springer
Herausgeber URL
http://dx.doi.org/10.3758/s13428-012-0281-2
Datum der Datenerfassung
2020
Datum, an dem der Datensatz öffentlich gemacht wurde
2020
Zugang
Public
Titel
Evaluating the robustness of repeated measures analyses : the case of small sample sizes and nonnormal data
Ausgabe der Zeitschrift
45

Datenquelle: METADATA.UB

Andere Metadatenquellen:
Autoren
  • Daniel Oberfeld
  • Thomas Franke
Autoren-URL
https://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=fis-test-1&SrcAuth=WosAPI&KeyUT=WOS:000328269200020&DestLinkType=FullRecord&DestApp=WOS_CPL
DOI
10.3758/s13428-012-0281-2
eISSN
1554-3528
Externe Identifier
  • Clarivate Analytics Document Solution ID: 269VA
  • PubMed Identifier: 23184532
ISSN
1554-351X
Ausgabe der Veröffentlichung
3
Zeitschrift
BEHAVIOR RESEARCH METHODS
Schlüsselwörter
  • Analysis of variance
  • Robustness
  • Nonnormality
  • Small sample settings
  • Repeated measurements
  • Correlated data
  • Multivariate
  • Mixed model analyses
  • Multilevel model
  • Simulation study
  • Type I error rate
  • Central limit theorem
  • Monte Carlo
Paginierung
792 - 812
Datum der Veröffentlichung
2013
Status
Published
Titel
Evaluating the robustness of repeated measures analyses: The case of small sample sizes and nonnormal data
Sub types
  • Article
Ausgabe der Zeitschrift
45

Datenquelle: Web of Science (Lite)

Autoren
  • Daniel Oberfeld
  • Thomas Franke
DOI
10.3758/s13428-012-0281-2
eISSN
1554-3528
Ausgabe der Veröffentlichung
3
Zeitschrift
Behavior Research Methods
Sprache
en
Online publication date
2012
Paginierung
792 - 812
Datum der Veröffentlichung
2013
Status
Published
Herausgeber
Springer Science and Business Media LLC
Herausgeber URL
http://dx.doi.org/10.3758/s13428-012-0281-2
Datum der Datenerfassung
2024
Titel
Evaluating the robustness of repeated measures analyses: The case of small sample sizes and nonnormal data
Ausgabe der Zeitschrift
45

Datenquelle: Crossref

Abstract
Repeated measures analyses of variance are the method of choice in many studies from experimental psychology and the neurosciences. Data from these fields are often characterized by small sample sizes, high numbers of factor levels of the within-subjects factor(s), and nonnormally distributed response variables such as response times. For a design with a single within-subjects factor, we investigated Type I error control in univariate tests with corrected degrees of freedom, the multivariate approach, and a mixed-model (multilevel) approach (SAS PROC MIXED) with Kenward-Roger's adjusted degrees of freedom. We simulated multivariate normal and nonnormal distributions with varied population variance-covariance structures (spherical and nonspherical), sample sizes (N), and numbers of factor levels (K). For normally distributed data, as expected, the univariate approach with Huynh-Feldt correction controlled the Type I error rate with only very few exceptions, even if samples sizes as low as three were combined with high numbers of factor levels. The multivariate approach also controlled the Type I error rate, but it requires N ≥ K. PROC MIXED often showed acceptable control of the Type I error rate for normal data, but it also produced several liberal or conservative results. For nonnormal data, all of the procedures showed clear deviations from the nominal Type I error rate in many conditions, even for sample sizes greater than 50. Thus, none of these approaches can be considered robust if the response variable is nonnormally distributed. The results indicate that both the variance heterogeneity and covariance heterogeneity of the population covariance matrices affect the error rates.
Addresses
  • Department of Psychology, Johannes Gutenberg-Universität, 55099 Mainz, Germany. oberfeld@uni-mainz.de
Autoren
  • Daniel Oberfeld
  • Thomas Franke
DOI
10.3758/s13428-012-0281-2
eISSN
1554-3528
Externe Identifier
  • PubMed Identifier: 23184532
Open access
false
ISSN
1554-351X
Ausgabe der Veröffentlichung
3
Zeitschrift
Behavior research methods
Schlüsselwörter
  • Humans
  • Analysis of Variance
  • Models, Statistical
  • Normal Distribution
  • Sample Size
  • Psychology, Experimental
  • Research Design
  • Male
Sprache
eng
Medium
Print
Paginierung
792 - 812
Datum der Veröffentlichung
2013
Status
Published
Datum der Datenerfassung
2012
Titel
Evaluating the robustness of repeated measures analyses: the case of small sample sizes and nonnormal data.
Sub types
  • Research Support, Non-U.S. Gov't
  • Evaluation Study
  • Journal Article
Ausgabe der Zeitschrift
45

Datenquelle: Europe PubMed Central

Abstract
Repeated measures analyses of variance are the method of choice in many studies from experimental psychology and the neurosciences. Data from these fields are often characterized by small sample sizes, high numbers of factor levels of the within-subjects factor(s), and nonnormally distributed response variables such as response times. For a design with a single within-subjects factor, we investigated Type I error control in univariate tests with corrected degrees of freedom, the multivariate approach, and a mixed-model (multilevel) approach (SAS PROC MIXED) with Kenward-Roger's adjusted degrees of freedom. We simulated multivariate normal and nonnormal distributions with varied population variance-covariance structures (spherical and nonspherical), sample sizes (N), and numbers of factor levels (K). For normally distributed data, as expected, the univariate approach with Huynh-Feldt correction controlled the Type I error rate with only very few exceptions, even if samples sizes as low as three were combined with high numbers of factor levels. The multivariate approach also controlled the Type I error rate, but it requires N ≥ K. PROC MIXED often showed acceptable control of the Type I error rate for normal data, but it also produced several liberal or conservative results. For nonnormal data, all of the procedures showed clear deviations from the nominal Type I error rate in many conditions, even for sample sizes greater than 50. Thus, none of these approaches can be considered robust if the response variable is nonnormally distributed. The results indicate that both the variance heterogeneity and covariance heterogeneity of the population covariance matrices affect the error rates.
Autoren
  • Daniel Oberfeld
  • Thomas Franke
Autoren-URL
https://www.ncbi.nlm.nih.gov/pubmed/23184532
DOI
10.3758/s13428-012-0281-2
eISSN
1554-3528
Ausgabe der Veröffentlichung
3
Zeitschrift
Behav Res Methods
Schlüsselwörter
  • Analysis of Variance
  • Humans
  • Male
  • Models, Statistical
  • Normal Distribution
  • Psychology, Experimental
  • Research Design
  • Sample Size
Sprache
eng
Country
United States
Paginierung
792 - 812
Datum der Veröffentlichung
2013
Status
Published
Datum, an dem der Datensatz öffentlich gemacht wurde
2013
Titel
Evaluating the robustness of repeated measures analyses: the case of small sample sizes and nonnormal data.
Sub types
  • Evaluation Study
  • Journal Article
  • Research Support, Non-U.S. Gov't
Ausgabe der Zeitschrift
45

Datenquelle: PubMed

Beziehungen:
Eigentum von

Werkzeuge