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

Maximum likelihood estimation of the kappa coefficient from bivariate logistic regression

M M Shoukri1, I U Mian

  • 1University of Guelph, Department of Population Medicine, Ontario, Canada.

Statistics in Medicine
|July 15, 1996
PubMed
Summary

This study introduces a new statistical method for estimating the kappa coefficient in binary ratings, accounting for patient and clinician influences. The developed method also aids in determining sample sizes for hypothesis testing in these scenarios.

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

Regional variation in organ donation in Saudi Arabia.

Transplantation proceedings·2014
Same author

Bivariate models for co-aggregation of dichotomous traits in twins.

Statistics in medicine·2006
Same author

Issues of cost and efficiency in the design of reliability studies.

Biometrics·2004
Same author

Efficiency considerations in the analysis of inter-observer agreement.

Biostatistics (Oxford, England)·2003
Same author

Air emissions from sour-gas processing plants and dairy-cattle reproduction in Alberta, Canada.

Preventive veterinary medicine·2003
Same author

Caffeine metabolism in premature infants.

Journal of clinical pharmacology·2001

Area of Science:

  • Biostatistics
  • Statistical Modeling
  • Medical Informatics

Background:

  • Inter-rater reliability is crucial for diagnostic accuracy.
  • Traditional kappa coefficient may be biased when ratings are influenced by patient or clinician factors.
  • Modeling dependent binary ratings requires advanced statistical approaches.

Purpose of the Study:

  • To develop a maximum likelihood estimator (MLE) for the kappa coefficient in 2x2 tables.
  • To account for patient and/or clinician effects on binary ratings.
  • To provide a method for sample size calculation for hypothesis testing of the kappa coefficient.

Main Methods:

  • Proposed a maximum likelihood estimator (MLE) for the kappa coefficient.
  • Modeled the logit of positive rating probability as a linear function of subject-specific and rater-specific covariates.

Related Experiment Videos

  • Utilized Monte Carlo simulations to assess bias and variance of the MLE in small to moderate samples.
  • Developed sample size calculation for detecting departures from a null hypothesis (H0: kappa = kappa 0).

    • Main Results:

    • The proposed MLE provides an estimate of the kappa coefficient under dependent binary ratings.
    • Monte Carlo simulations evaluated the performance of the MLE regarding bias and variance.
    • A method for sample size determination was established for hypothesis testing.

    Conclusions:

    • The developed MLE is suitable for estimating kappa coefficients when binary ratings are influenced by patient and/or clinician effects.
    • The study provides a framework for sample size calculation in such scenarios.
    • This work contributes to more accurate reliability assessments in medical diagnostics.