Jove
Visualize
Contact Us

Related Experiment Videos

Bias-corrected maximum likelihood estimator of the intraclass correlation parameter for binary data.

Krishna K Saha1, Sudhir R Paul

  • 1Department of Mathematics and Statistics, University of Windsor, 401 Sunset, Windsor, Ont., Canada N9B 3P4. smjp@uwindsor.ca

Statistics in Medicine
|July 12, 2005
PubMed
Summary

This study introduces a bias-corrected maximum likelihood (BCML) estimator for the intraclass correlation parameter in extended beta-binomial models. The BCML estimator demonstrates improved bias and efficiency compared to existing methods, enhancing statistical analysis accuracy.

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

Confidence intervals for the difference in the success rates of two treatments in the analysis of correlated binary responses.

Biometrical journal. Biometrische Zeitschrift·2019
Same author

Confidence intervals for the common intraclass correlation in the analysis of clustered binary responses.

Journal of biopharmaceutical statistics·2017
Same author

A Comparison of Some Approximate Confidence Intervals for a Single Proportion for Clustered Binary Outcome Data.

The international journal of biostatistics·2015
Same author

Weighted profile likelihood-based confidence interval for the difference between two proportions with paired binomial data.

Statistics in medicine·2014
Same author

Semiparametric approach for non-monotone missing covariates in a parametric regression model.

Biometrics·2014
Same author

Inference concerning a common dispersion of several treatment groups in the analysis of over/underdispersed count data.

Biometrical journal. Biometrische Zeitschrift·2014
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:

  • Biostatistics
  • Statistical Modeling
  • Correlations

Background:

  • The extended beta-binomial model is commonly used for over/under-dispersed proportions.
  • Maximum likelihood estimation (MLE) of the dispersion parameter (phi) can be biased with small sample sizes or low Fisher information.
  • Accurate estimation of the intraclass correlation parameter is crucial for reliable statistical inference.

Purpose of the Study:

  • To develop and evaluate a bias-corrected maximum likelihood (BCML) estimator for the intraclass correlation parameter.
  • To compare the performance of the BCML estimator against existing methods, including MLE, Q(2), and DEQL.
  • To assess the practical utility of the BCML estimator using real-world toxicological and medical datasets.

Main Methods:

  • Derivation of a bias-corrected maximum likelihood (BCML) estimator for the intraclass correlation parameter.

Related Experiment Videos

  • Monte Carlo simulations to compare bias and efficiency of BCML with MLE, Q(2), and DEQL estimators.
  • Application of the estimators to toxicological data and medical data on chromosomal abnormalities.

    • Main Results:

    • The BCML estimator exhibited superior bias and efficiency properties in most simulated scenarios.
    • Analyses of real-world data showed significant improvements in standard errors for BCML estimates.
    • The BCML estimator provides a more reliable alternative for estimating intraclass correlation in extended beta-binomial models.

    Conclusions:

    • The proposed BCML estimator offers a statistically robust improvement over existing methods for estimating the intraclass correlation parameter.
    • BCML provides more accurate and efficient estimates, leading to improved standard errors in practical applications.
    • This bias-corrected approach enhances the reliability of analyses involving over/under-dispersed proportional data.