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

Tests for homogeneity of the risk difference when data are sparse

S R Lipsitz1, K B Dear, N M Laird

  • 1Department of Biostatistics, Harvard School of Public Health, Boston, Massachusetts 02115, USA. lipsitz@biostat.harvard.edu

Biometrics
|April 17, 1998
PubMed
Summary

New test statistics address sparse data issues in risk difference homogeneity testing. Simulations show the proposed methods outperform the common weighted least squares statistic, offering more reliable Type I error rates.

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

Sick leave due to SARS-CoV-2 infection.

Occupational medicine (Oxford, England)·2023
Same author

Geographical variation of COVID-19 vaccination coverage, ethnic diversity and population composition in Flanders.

Vaccine: X·2022
Same author

Aligning Estimators With Estimands in Clinical Trials: Putting the ICH E9(R1) Guidelines Into Practice.

Therapeutic innovation & regulatory science·2020
Same author

Generalized pairwise comparison methods to analyze (non)prioritized composite endpoints.

Statistics in medicine·2019
Same author

Reducing surgical mortality in Scotland by use of the WHO Surgical Safety Checklist.

The British journal of surgery·2019
Same author

Recent Developments in the Prevention and Treatment of Missing Data.

Therapeutic innovation & regulatory science·2018

Area of Science:

  • Biostatistics
  • Epidemiology
  • Statistical Methods

Background:

  • Testing for homogeneity of risk difference across multiple 2x2 tables is crucial in meta-analysis.
  • The standard weighted least squares statistic may exhibit anticonservative Type I error rates with sparse data.
  • Sparse data is common in certain epidemiological studies, necessitating robust statistical approaches.

Purpose of the Study:

  • To propose novel test statistics for assessing the homogeneity of risk difference in sparse 2x2 table data.
  • To evaluate the performance of these new statistics against the conventional weighted least squares method.
  • To provide a more reliable statistical tool for meta-analyses involving sparse data.

Main Methods:

  • Development of new test statistics specifically designed for sparse 2x2 contingency tables.

Related Experiment Videos

  • Comparative analysis using Monte Carlo simulations to assess Type I error rates.
  • Evaluation of the weighted least squares statistic under conditions of data sparsity.

    • Main Results:

    • The weighted least squares statistic demonstrated the most anticonservative Type I error rates among all tested statistics.
    • Simulations indicated that the proposed test statistics maintained more appropriate Type I error rates.
    • The new statistics showed superior performance in controlling false positive rates when data are sparse.

    Conclusions:

    • The proposed test statistics are recommended as a more reliable alternative to the weighted least squares statistic for homogeneity testing with sparse data.
    • Adoption of these new methods can improve the accuracy of meta-analyses and risk difference evaluations in challenging data scenarios.
    • Further research could explore the power of these proposed statistics in various sparse data settings.