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

Exposure misclassification: bias in category specific Poisson regression coefficients.

M B Veierød1, P Laake

  • 1Section of Medical Statistics, University of Oslo, P.O. Box 1122, Blindern, 0317 Oslo, Norway. marit.veierod@basalmed.uio.no

Statistics in Medicine
|March 10, 2001
PubMed
Summary
This summary is machine-generated.

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

Menopausal hormone therapy and risk of melanoma: a population-based cohort study.

The British journal of dermatology·2021
Same author

Body weight, body composition and survival after 1 year: follow-up of a nutritional intervention trial in allo-HSCT recipients.

Bone marrow transplantation·2019
Same author

Reproductive factors and risk of melanoma: a population-based cohort study.

The British journal of dermatology·2019
Same author

Why a randomized melanoma screening trial is not a good idea.

The British journal of dermatology·2018
Same author

Anthropometric factors and Breslow thickness: prospective data on 2570 cases of cutaneous melanoma in the population-based Janus Cohort.

The British journal of dermatology·2018
Same author

Ultraviolet exposure from indoor tanning devices: a systematic review.

The British journal of dermatology·2016

Exposure misclassification in epidemiologic studies can bias results. This study presents a method to assess bias in Poisson regression coefficients for categorical exposure, revealing complex bias patterns.

Area of Science:

  • Epidemiology
  • Biostatistics
  • Statistical Modeling

Background:

  • Exposure misclassification is a common issue in epidemiologic studies.
  • Misclassification can introduce bias into effect estimates, making interpretation challenging.
  • The direction and magnitude of this bias are often unpredictable.

Purpose of the Study:

  • To develop a simple method for assessing bias in Poisson regression coefficients.
  • To analyze bias specifically for categorical exposure variables with misclassification.
  • To provide a framework for understanding the impact of exposure misclassification on regression models.

Main Methods:

  • Derivation of expressions relating naive (error-prone) and true coefficients.
  • Analysis under assumptions of error-free covariates and independence of misclassification probabilities.

Related Experiment Videos

  • Application to Poisson regression models for categorical exposure.
  • Illustration using simulated scenarios of exposure-disease associations and misclassification patterns.

    • Main Results:

    • Bias in naive coefficients depends on all true category-specific coefficients and misclassification probabilities.
    • Misclassification of exposure does not bias coefficients for perfectly measured covariates.
    • The bias can be complex, varying in magnitude and direction across exposure categories.
    • Intuitive characterization of bias is often difficult.

    Conclusions:

    • The developed method aids in assessing exposure misclassification bias in Poisson regression.
    • Understanding bias is crucial for accurate interpretation of epidemiologic findings.
    • The study highlights the intricate nature of bias induced by misclassified categorical exposures.