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

Type I and Type II error under random-effects misspecification in generalized linear mixed models.

Saskia Litière1, Ariel Alonso, Geert Molenberghs

  • 1Center for Statistics, Hasselt University, Agoralaan, Building D, B3590 Diepenbeek, Belgium. saskia.litiere@uhasselt.be

Biometrics
|April 12, 2007
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

Somatostatin receptor PET response assessment framework for patients with neuroendocrine tumours (V1.0): a modified Delphi consensus from the European Neuroendocrine Tumor Society (endorsed by EANM and NANETS).

The Lancet. Oncology·2026
Same author

Evaluation of surrogate endpoints for survival outcomes using the surrogate package in R.

Computer methods and programs in biomedicine·2026
Same author

Integrating oral health screening into general practice: validation study of the Oral Health Screener.

Scientific reports·2026
Same author

Time-Scale Target Parameters and Two-Step Estimation in Longitudinal Trials for Progressive Diseases.

Statistics in medicine·2026
Same author

European Organisation for Research and Treatment of Cancer (EORTC) 2025 Program for Early Career Investigators: Advancing Ideas into Clinical Research.

European journal of cancer (Oxford, England : 1990)·2026
Same author

Handling Missing Data in Participants with Baseline but No Post-Baseline Data.

Pharmaceutical statistics·2026

Misspecifying random-effects distributions in generalized linear mixed models (GLMMs) can inflate type I errors and affect test power. However, some significant effects remain reliable even with misspecification.

Area of Science:

  • Statistics
  • Biostatistics
  • Longitudinal Data Analysis

Background:

  • Generalized linear mixed models (GLMMs) are widely used for non-Gaussian longitudinal data.
  • Maximum likelihood estimation in GLMMs assumes correct model specification.
  • GLMM results may lack robustness when assumptions are violated.

Purpose of the Study:

  • To investigate the impact of misspecifying random-effects distributions in logistic random-intercept GLMMs.
  • To assess the effects on Type I and Type II errors for mean structure tests.
  • To evaluate the reliability of GLMM results under distributional misspecification.

Main Methods:

  • Utilized simulation studies with a logistic random-intercept model.
  • Analyzed the influence of varying random-effects distributions on statistical tests.

Related Experiment Videos

  • Derived a theoretical result concerning the consistency of fixed-effects estimators.

    • Main Results:

    • Misspecification of random-effects distribution can increase or decrease test power.
    • Type I error rates can be considerably inflated due to misspecification.
    • A theoretical finding indicates consistent estimation of zero for certain fixed effects, preserving reliability.

    Conclusions:

    • Random-effects distribution misspecification poses a risk to the validity of GLMM analyses.
    • The impact on statistical power is dependent on the specific distribution shape.
    • Under specific conditions, significant GLMM results can remain trustworthy despite distributional misspecification.