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

The bias due to incomplete matching.

P R Rosenbaum, D B Rubin

    Biometrics
    |March 1, 1985
    PubMed
    Summary

    Matching methods in observational studies can introduce bias. This study quantizes bias from incomplete and inexact matching, showing a multivariate nearest available matching algorithm effectively minimizes bias.

    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

    Counternull sets in randomized experiments.

    The American statistician·2025
    Same author

    Matching One Sample According to Two Criteria in Observational Studies.

    Journal of the American Statistical Association·2023
    Same author

    High-dimensional randomization-based inference capitalizing on classical design and modern computing.

    Behaviormetrika·2023
    Same author

    The importance of having a conceptual stage when reporting non-randomized studies.

    Biostatistics & epidemiology·2021
    Same author

    Heterogeneous ozone effects on the DNA methylome of bronchial cells observed in a crossover study.

    Scientific reports·2020
    Same author

    When possible, report a Fisher-exact <i>P</i> value and display its underlying null randomization distribution.

    Proceedings of the National Academy of Sciences of the United States of America·2020

    Area of Science:

    • Biostatistics
    • Epidemiology
    • Observational Studies

    Background:

    • Observational studies are crucial for estimating treatment effects.
    • Matching is a common technique to create comparable control groups.
    • A trade-off exists between matching all treated units and achieving high similarity.

    Purpose of the Study:

    • To derive expressions for bias in matching methods.
    • To quantify bias from incomplete and inexact matching.
    • To evaluate bias mitigation strategies.

    Main Methods:

    • Derivation of mathematical expressions for matching bias.
    • Analysis of bias due to incomplete matching (failure to match all units).
    • Analysis of bias due to inexact matching (imperfect similarity).

    Main Results:

    • Bias from incomplete matching can be substantial.
    • Bias from inexact matching is typically smaller than from incomplete matching.
    • A multivariate nearest available matching algorithm effectively reduces bias from incomplete matching.

    Conclusions:

    • Incomplete matching can severely bias treatment effect estimates.
    • Multivariate nearest available matching is a robust method to minimize bias.
    • Careful selection of matching algorithms is essential for reliable observational study results.

    Related Experiment Videos