Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

A note on testing for homogeneity among effect sizes sharing a common control group.

Samantha R Cook1

  • 1Harvard University, Cambridge, MA, USA. cook@stat.columbia.edu

Psychological Methods
|December 16, 2004
PubMed
Summary

This study offers a more exact covariance matrix for correlated effect sizes in meta-analysis, improving homogeneity testing. The method enhances statistical accuracy for correlated effect sizes in research synthesis.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Mineralised organic remains from cesspits at the Roman town of Silchester: processes and preservation.

Spectrochimica acta. Part A, Molecular and biomolecular spectroscopy·2008
Same author

Analysis of ochres from Clearwell Caves: the role of particle size in determining colour.

Spectrochimica acta. Part A, Molecular and biomolecular spectroscopy·2004
Same author

Antisaccade performance in biological relatives of schizophrenia patients: a meta-analysis.

Schizophrenia research·2004
See all related articles

Area of Science:

  • Statistics
  • Meta-Analysis
  • Psychiatry Research

Background:

  • Homogeneity testing is crucial in meta-analysis for synthesizing correlated effect sizes.
  • Existing methods rely on large-sample approximations for the covariance matrix of correlated effect sizes.
  • A specific scenario involves a single control group compared with multiple treatment groups, leading to correlated effect sizes.

Purpose of the Study:

  • To derive a more exact expression for the covariance matrix of correlated effect sizes.
  • To provide an improved method for testing homogeneity among correlated effect sizes without relying on large-sample assumptions.
  • To apply the enhanced method to schizophrenia research data.

Main Methods:

  • Derivation of an exact covariance matrix expression for correlated effect sizes under normality assumptions.

Related Experiment Videos

  • Estimation of the correlation between effect sizes.
  • Application of the standard Q statistic for correlated effect sizes using the derived covariance matrix.

    • Main Results:

    • A more exact formula for the covariance matrix of correlated effect sizes was established.
    • The enhanced method allows for more precise homogeneity testing, especially when sample sizes are not large.
    • The methodology was successfully demonstrated using data from schizophrenia studies.

    Conclusions:

    • The presented exact covariance matrix improves the accuracy of homogeneity tests for correlated effect sizes.
    • This statistical advancement is particularly valuable in meta-analyses with complex group comparisons.
    • The findings offer a more robust approach to synthesizing evidence in fields like schizophrenia research.