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

Berkson's bias revisited.

E Peritz

    Journal of Chronic Diseases
    |January 1, 1984
    PubMed
    Summary

    Berkson's bias, a form of Simpson's paradox, distorts associations in hospitalized patients. Introducing an independence assumption shows this bias affects selected subjects, decreasing odds-ratios and allowing for temporal analysis.

    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

    Mortality from lymphomas in Israel, 1950-71: the possible role of environmental factors.

    International journal of epidemiology·2013
    Same author

    Mortality differentials among Israeli men.

    American journal of public health·1999
    Same author

    The Israel Longitudinal Mortality Study--differential mortality in Israel 1983-1992: objectives, materials, methods and preliminary results.

    Israel journal of medical sciences·1998
    Same author

    Re: "Multiple comparisons and related issues in the interpretation of epidemiologic data".

    American journal of epidemiology·1997
    Same author

    Potential demography: a second look.

    European journal of population = Revue europeenne de demographie·1996
    Same author

    The Turlock vitamin-mineral supplementation trial: a statistical reanalysis.

    Journal of biosocial science·1994

    Area of Science:

    • Epidemiology
    • Biostatistics

    Background:

    • Berkson's bias, a statistical phenomenon, can distort associations observed in clinical research.
    • It is often considered a specific instance of Simpson's paradox, particularly in the context of hospitalized patient populations.

    Purpose of the Study:

    • To clarify the nature of Berkson's bias and its relationship with Simpson's paradox.
    • To analyze the impact of Berkson's independence assumption on study subjects and effect measures.
    • To explore the applicability of the Berkson's bias model across different study designs and its potential for temporal analysis.

    Main Methods:

    • The study employs a theoretical modeling approach to define and analyze Berkson's bias.
    • It differentiates between a general interpretation and a specific interpretation involving an independence assumption.
    • The model's validity is discussed for case-control, prospective, and other observational study designs.

    Main Results:

    • Berkson's bias is presented as a special case of Simpson's paradox when applied loosely to hospitalized patients.
    • When Berkson's independence assumption is applied, the bias specifically impacts selected subjects, not those excluded.
    • This assumption tends to decrease the observed odds-ratio.

    Conclusions:

    • The model for Berkson's bias is broadly applicable across various epidemiological study designs.
    • Incorporating a time dimension enables the investigation of dynamic changes in Berkson's bias over time.
    • Understanding Berkson's bias is crucial for accurate interpretation of associations in observational research.

    Related Experiment Videos