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

Methods to account for attrition in longitudinal data: do they work? A simulation study.

Vicki L Kristman1, Michael Manno, Pierre Côté

  • 1Institute for Work and Health, Toronto, Ontario, Canada. vkristman@iwh.on.ca

European Journal of Epidemiology
|September 10, 2005
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

The association between low back pain and sleep quality in adults 60 years and older: a systematic review.

Chiropractic & manual therapies·2026
Same author

Structural barriers and facilitators of accessing rehabilitation among adults aged 50 years and older with low back pain: a scoping review of the literature.

BMC health services research·2026
Same author

Set of Cochrane Overviews of Systematic Reviews Show the Limits of Current Evidence on Health Policy, Systems, and Services Research in Rehabilitation.

American journal of physical medicine & rehabilitation·2026
Same author

Assessing Cultural Competence Among Canadian Physiotherapists: A Qualitative Analysis of a Cross-Sectional Survey, Part 2.

Physiotherapy Canada. Physiotherapie Canada·2026
Same author

Assessing Diversity and Cultural Competency Among Canadian Physiotherapists: A Cross-Sectional Survey, Part 1.

Physiotherapy Canada. Physiotherapie Canada·2026
Same author

Delivery arrangements for rehabilitation services in health systems: an overview of systematic reviews.

The Cochrane database of systematic reviews·2026

Attrition bias in cohort studies is a concern. Standard methods like regression imputation, individual weighting, and multiple imputation may not effectively reduce bias when data are missing not at random (MNAR).

Area of Science:

  • Epidemiology
  • Biostatistics

Background:

  • Cohort studies are susceptible to internal validity threats from participant attrition.
  • Existing statistical methods like imputation and weighting aim to mitigate attrition bias, but their effectiveness is not fully established.

Purpose of the Study:

  • To evaluate the efficacy of regression imputation, individual weighting, and multiple imputation in reducing attrition bias in cohort studies.
  • To compare these methods against complete case analysis under various missing data scenarios.

Main Methods:

  • A simulated cohort of 300 subjects was analyzed using 500 replications.
  • Attrition rates of 10%, 25%, and 40% were generated under missing completely at random (MCAR), missing at random (MAR), and missing not at random (MNAR) mechanisms.
  • A logistic regression model with a binary exposure and two confounders was employed.

Related Experiment Videos

Main Results:

  • Under MCAR and MAR conditions, complete case analysis yielded results comparable to or better than imputation and weighting methods across all attrition levels.
  • When data were MNAR, none of the tested methods produced unbiased odds ratio estimates at 25% or 40% attrition.

Conclusions:

  • Complete case analysis is often as valid as advanced methods when data are missing completely at random or missing at random.
  • Imputation and weighting methods demonstrate limited effectiveness in correcting attrition bias when data are missing not at random.