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

Linear regression analysis of binary response data with mixed covariates - a simulation study.

P D Wilson, C L Meinert

    Controlled Clinical Trials
    |March 1, 1984
    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

    Maternal birth trauma and its consequences: time to raise awareness.

    International urogynecology journal·2021
    Same author

    Randomized Trial of 2 Self-Titrated Oral Appliances for Airway Management.

    Journal of dental research·2020
    Same author

    The IUGA Obstetric Pelvic Floor Trauma Special Interest Group's response to Gachon B et al/ EBCOG re: Should we advise women that pre-labor caesarean section prevents pelvic floor dysfunction?

    European journal of obstetrics, gynecology, and reproductive biology·2020
    Same author

    Insurance status and mortality among patients with AIDS.

    HIV medicine·2017
    Same author

    Authors' comment: A report from #BlueJC: 'How to manage incontinence after childbirth?'.

    BJOG : an international journal of obstetrics and gynaecology·2014
    Same author

    Diffusion-weighted and multiphase contrast-enhanced MRI as surrogate markers of response to neoadjuvant sunitinib in metastatic renal cell carcinoma.

    British journal of cancer·2013
    Same journal

    On the generation and ownership of alpha in medical studies.

    Controlled clinical trials·2004
    Same journal

    An analysis of the effect of funding source in randomized clinical trials of second generation antipsychotics for the treatment of schizophrenia.

    Controlled clinical trials·2004
    Same journal

    Symptom recording in a randomised clinical trial: paper diaries vs. electronic or telephone data capture.

    Controlled clinical trials·2004
    Same journal

    Statistical comparison of random allocation methods in cancer clinical trials.

    Controlled clinical trials·2004
    Same journal

    Analyzing bronchodilation with emphasis on disease type, age and sex.

    Controlled clinical trials·2004
    Same journal

    Geographic variability in patient characteristics, treatment and outcome in an International Trial of Magnesium in acute myocardial infarction.

    Controlled clinical trials·2004
    See all related articles

    Linear regression analysis for binary outcomes with mixed covariates can lead to inaccurate Type I error rates. This occurs when covariate distributions vary across treatment groups, deviating from Gaussian assumptions.

    Area of Science:

    • Biostatistics
    • Clinical Trial Methodology
    • Statistical Modeling

    Background:

    • Clinical trials often involve binary outcomes and require covariate adjustment for accurate treatment comparisons.
    • Mixed covariates (binary and continuous/Gaussian) present analytical challenges, with logistic models being theoretically appropriate.
    • Historically, both linear and logistic regression have been applied, potentially leading to discrepancies.

    Purpose of the Study:

    • To investigate the Type I error rate of linear regression tests for treatment comparisons in binary logistic models with mixed covariates.
    • To assess the impact of covariate distribution and effect size on the accuracy of linear regression analyses.
    • To identify potential issues arising from applying linear regression to data that follows a logistic model.

    Main Methods:

    Related Experiment Videos

    • Computer simulations were employed to generate data reflecting a logistic model with mixed binary and Gaussian covariates.
    • Linear regression analysis tests were applied to compare treatment groups.
    • The Type I error level was evaluated under varying covariate distributions and effect magnitudes.

    Main Results:

    • The true Type I error rate of linear regression tests was found to be dependent on the distribution of covariates across treatment groups.
    • The magnitude of covariate effects significantly influenced the observed Type I error rate.
    • Deviations from the assumed Gaussian distribution led to important differences in Type I error rates compared to theoretical expectations.

    Conclusions:

    • Linear regression analysis is not always appropriate for binary outcomes with mixed covariates, as it can inflate or deflate Type I error rates.
    • The accuracy of linear regression is compromised when covariate distributions differ between treatment groups or when covariate effects are substantial.
    • Recommendations are needed for the development of statistical software packages to handle such complex data structures appropriately.