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

Complete imputation of missing repeated categorical data: one-sample applications.

Colin P West1, Jeffrey D Dawson

  • 1Mayo Graduate School of Medicine, Rochester, MN 55905, USA. west.colin@mayo.edu

Statistics in Medicine
|January 10, 2002
PubMed
Summary

This study introduces a new method for handling missing categorical data in longitudinal studies, improving analysis when missingness depends on outcomes. It offers a more comprehensive way to interpret study findings for better statistical significance.

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

Assessment of psychological and physiological responses to auditory and visual stimuli during recovery from acute stress induction.

Physiology international·2026
Same author

Occupational distress factors associated with suicidal ideation among physicians and comparison with other US workers.

Academic medicine : journal of the Association of American Medical Colleges·2026
Same author

Crisis-Related Self-Reported symptoms of posttraumatic stress disorder in graduate medical education trainees: Results from a Multi-Institutional study.

Academic medicine : journal of the Association of American Medical Colleges·2026
Same author

Associations of spousal support with career attrition in academic medicine: a multi-specialty cross-sectional survey study.

BMJ leader·2026
Same author

Associations and Differences Between Occupational Burnout and Depression in Large Studies of U.S. Physicians.

Academic medicine : journal of the Association of American Medical Colleges·2026
Same author

The validation of measured and self-reported sleep duration and perceived sleep quality: an empirical study with three generations of smartwatches.

Frontiers in psychiatry·2026

Area of Science:

  • Statistics
  • Biostatistics
  • Longitudinal Data Analysis

Background:

  • Longitudinal studies with repeated measures frequently encounter non-response.
  • Existing methods for handling missing data are often inadequate, particularly when missingness is outcome-dependent.

Purpose of the Study:

  • To present a novel approach for analyzing incomplete categorical data in repeated measures settings.
  • To develop a methodology where missing data can depend on other observed outcomes.
  • To enable a broader examination of study findings by considering all potential test statistics.

Main Methods:

  • Generating all possible sets of missing values to create a set of potential complete datasets.
  • Weighting each dataset based on defined assumptions.

Related Experiment Videos

  • Applying statistical tests to each dataset and combining results for overall significance.
  • Utilizing the Expectation-Maximization (EM) algorithm and a Bayesian prior.

    • Main Results:

    • The proposed methodology effectively handles missing categorical data in longitudinal studies.
    • It allows for missing data mechanisms related to observed outcomes.
    • Demonstrated effectiveness in a one-sample case, outperforming complete-case and available-case analyses.

    Conclusions:

    • The presented approach offers a robust solution for incomplete categorical data in repeated measures.
    • It provides a more comprehensive framework for interpreting statistical findings in the presence of missing data.
    • This method enhances the reliability of analyses in longitudinal research where non-response is common.