Jove
Visualize
Contact Us

Related Experiment Videos

Adjusting for attrition in event-history analysis.

D H Hill

    Sociological Methodology
    |January 1, 1997
    PubMed
    Summary

    Weighting methods significantly impact event-history models, especially when accounting for panel attrition in divorce studies. Incorrectly handling attrition can lead to misclassified divorces as panel attrition in surveys.

    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

    Low thermal inertia of carbonaceous asteroid Bennu driven by cracks observed in returned samples.

    Nature communications·2026
    Same author

    An evaporite sequence from ancient brine recorded in Bennu samples.

    Nature·2025
    Same author

    Petrography and bulk chemistry of Martian orthopyroxenite ALH84001: implications for the origin of secondary carbonates.

    Geochimica et cosmochimica acta·1997
    Same author

    A general overview of the anatomy of the neurohypophysis.

    Annals of the New York Academy of Sciences·1993
    Same author

    Metachronous colonic lymphomas complicating chronic ulcerative colitis.

    Abdominal imaging·1993
    Same author

    Competing hazards with shared unmeasured risk factors.

    Sociological methodology·1993
    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

    Area of Science:

    • Sociology
    • Statistics
    • Demography

    Background:

    • Panel attrition, where participants drop out of a study over time, poses a significant challenge in longitudinal research.
    • Event-history models are sensitive to how sampling weights are adjusted to account for participant attrition.
    • The Survey of Income and Program Participation (SIPP) is a key source for U.S. household economic data, making attrition a critical issue.

    Purpose of the Study:

    • To investigate the impact of different weighting procedures on event-history models concerning panel attrition.
    • To compare the effectiveness of initial selection probability weights, panel weights, and attrition-adjusted weights in a divorce model.
    • To assess whether attrition is misclassified as divorce in the SIPP data.

    Main Methods:

    Keywords:
    AmericasDeveloped CountriesDivorceError SourcesMeasurementMethodological StudiesModels, TheoreticalNorth AmericaNorthern AmericaNuptialityResearch MethodologySampling StudiesStudiesSurveysUnited States

    Related Experiment Videos

  • Utilized event-history models to analyze divorce data from the 1986 U.S. Survey of Income and Program Participation (SIPP).
  • Compared three distinct weighting procedures: initial selection probability weights, 1986 SIPP panel weights, and monthly attrition-adjusted weights.
  • Contrasted weighted estimates with those from a structural model treating attrition as an error-correlated competing alternative to divorce.

    • Main Results:

    • Different weighting procedures yield varying estimates in event-history models when addressing panel attrition.
    • A notable finding is that divorces within the SIPP dataset are frequently recorded as instances of panel attrition.
    • The choice of weighting method can substantially alter conclusions regarding divorce rates and patterns.

    Conclusions:

    • Careful consideration and appropriate adjustment of sampling weights are crucial for accurate analysis of event-history data with attrition.
    • The misclassification of divorces as attrition in the SIPP highlights potential data quality issues and biases.
    • Future research should focus on robust methods for distinguishing between true attrition and substantive events like divorce.