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 analysis for heteroscedastic one-way random effects ANOVA in interlaboratory studies.

M G Vangel1, A L Rukhin

  • 1Statistical Engineering Division, National Institute of Standards and Technology, Gaithersburg, Maryland 20899-0001, USA. vangel@cam.nist.gov

Biometrics
|April 25, 2001
PubMed
Summary

This study introduces a new method for analyzing measurements from multiple groups, simplifying calculations and improving the interpretation of results for random-effects ANOVA models with unequal variances.

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

An Approach to Combining Results From Multiple Methods Motivated by the ISO GUM.

Journal of research of the National Institute of Standards and Technology·2016
Same author

Behavioral inhibition in childhood predicts smaller hippocampal volume in adolescent offspring of parents with panic disorder.

Translational psychiatry·2015
Same author

Comparison of CE-FDG-PET/CT with CE-FDG-PET/MR in the evaluation of osseous metastases in breast cancer patients.

British journal of cancer·2015
Same author

Imaging appearance of the lateral rectus-superior rectus band in 100 consecutive patients without strabismus.

AJNR. American journal of neuroradiology·2014
Same author

Neuroanatomic correlates of stroke-related myocardial injury.

Neurology·2006
Same author

Case clustering in infective endocarditis: the role of availability bias.

Clinical microbiology and infection : the official publication of the European Society of Clinical Microbiology and Infectious Diseases·2005

Area of Science:

  • Statistics
  • Biometry
  • Experimental Design

Background:

  • Analyzing multiple groups of measurements on the same quantity is a common statistical challenge.
  • Existing methods, like those by Cochran, often involve complex calculations for unbalanced ANOVA with unequal variances.

Purpose of the Study:

  • To present a simplified maximum likelihood analysis for one-way unbalanced random-effects ANOVA with unequal within-group variances.
  • To improve the identification and interpretation of likelihood multimodality.
  • To provide approximate confidence regions for the mean and between-group variance using a Bayesian approach.

Main Methods:

  • Formulation as a one-way unbalanced random-effects ANOVA.
  • Reparametrization of the likelihood function.

Related Experiment Videos

  • Application of a non-informative-prior Bayesian approach.

    • Main Results:

    • Simplified computations for maximum likelihood analysis.
    • Enhanced identification and interpretation of likelihood multimodality.
    • Approximate confidence regions for key parameters (mean and between-group variance).

    Conclusions:

    • The proposed reparametrization offers computational advantages.
    • The method facilitates a clearer understanding of the likelihood's behavior.
    • Bayesian inference provides practical confidence regions for model parameters.