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

A robust approach to t linear mixed models applied to multiple sclerosis data.

Tsung I Lin1, Jack C Lee

  • 1Department of Applied Mathematics, National Chung Hsing University, Taichung 402, Taiwan. tilin@amath.nchu.edu.tw

Statistics in Medicine
|October 13, 2005
PubMed
Summary

This study introduces a robust statistical model for longitudinal data, enhancing linear mixed models with a multivariate t-distribution to better handle auto-correlated errors in clinical trials.

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

Design and application of synthetic 17B-HSD13 substrates reveals preserved catalytic activity of protective human variants.

Nature communications·2025
Same author

Enhancing simulation feasibility for multi-layer 2D MoS<sub>2</sub> RRAM devices: reliability performance learnings from a passive network model.

Physical chemistry chemical physics : PCCP·2024
Same author

Post-annealing optimization of the heteroepitaxial La-doped SrSnO<sub>3</sub> integrated on silicon <i>via</i> ALD.

Nanoscale·2023
Same author

Resistance state evolution under constant electric stress on a MoS<sub>2</sub> non-volatile resistive switching device.

RSC advances·2022
Same author

Wafer-Scalable Single-Layer Amorphous Molybdenum Trioxide.

ACS nano·2022
Same author

Wafer-Scale Synthesis of WS<sub>2</sub> Films with In Situ Controllable p-Type Doping by Atomic Layer Deposition.

Research (Washington, D.C.)·2021

Area of Science:

  • Statistics
  • Biostatistics
  • Longitudinal Data Analysis

Background:

  • Longitudinal data analysis requires robust models to account for time-dependent correlations.
  • Traditional linear mixed models may not adequately capture complex error structures in repeated measures.
  • Autocorrelation in within-subject errors is a common challenge in longitudinal studies.

Purpose of the Study:

  • To extend linear mixed models using the multivariate t-distribution for robust analysis of longitudinal data.
  • To develop methods for detecting and modeling autocorrelation in within-subject errors.
  • To provide tools for predicting future responses in longitudinal studies.

Main Methods:

  • Utilized a multivariate t-distribution to extend linear mixed models.

Related Experiment Videos

  • Employed a first-order autoregressive (AR(1)) dependence structure for within-subject errors.
  • Derived a score test statistic for autocorrelation and developed a maximum likelihood estimation (MLE) procedure.

    • Main Results:

    • The proposed extension provides a robust framework for longitudinal data with potential outliers or heavy tails.
    • A score test effectively detects autocorrelation in within-subject errors.
    • The MLE procedure yields standard errors and enables prediction of future responses.

    Conclusions:

    • The robust extension of linear mixed models offers improved performance for longitudinal data with auto-correlated errors.
    • The developed statistical tests and estimation procedures are valuable for analyzing complex longitudinal datasets.
    • The methodology is applicable to real-world clinical trial data, as demonstrated in a multiple sclerosis study.