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

Correcting covariate-dependent measurement error with non-zero mean.

Nabila Parveen1, Erica Moodie1, Bluma Brenner2

  • 1Department of Epidemiology, Biostatistics and Occupational Health, McGill University, Montreal, Quebec, Canada.

Statistics in Medicine
|April 11, 2017
PubMed
Summary

This study introduces a new statistical method to reduce bias in measurement error, especially when errors depend on other variables. The simulation-extrapolation extension performs better than existing methods for analyzing HIV transmission data.

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

It's about time: The association between abacavir and cardiovascular disease.

Antiviral therapy·2026
Same author

Estimating optimal dynamic treatment regimes with Gaussian process emulation.

Biometrics·2026
Same author

Joint mixed-effects models for causal inference in clustered network-based observational studies.

Statistical methods in medical research·2025
Same author

Causal mediation analysis with two mediators: A comprehensive guide to estimating total and natural effects across various multiple mediators setups.

Psychological methods·2025
Same author

A Framework for Generating Realistic Synthetic Tabular Data in a Randomized Controlled Trial Setting.

Statistics in medicine·2025
Same author

Sparse 2-stage Bayesian meta-analysis for individualized treatments.

Biometrics·2025

Area of Science:

  • Biostatistics
  • Epidemiology
  • Statistical Modeling

Background:

  • Measurement error in covariates can bias statistical analyses, particularly in epidemiological studies.
  • In HIV research, measurement error distribution may depend on other covariates like HIV status, complicating analysis.
  • Traditional methods for handling measurement error often require validation data or repeated measurements, which are not always feasible.

Purpose of the Study:

  • To propose an extension of the simulation-extrapolation (SIMEX) method to address complex measurement error structures.
  • To develop a technique that reduces bias without needing validation data or repeated measurements.
  • To accommodate measurement error distributions that vary with error-free covariates.

Main Methods:

  • An extension of the simulation-extrapolation (SIMEX) estimation technique was developed.
Keywords:
HIVbiasmeasurement errorsimulation-extrapolation

Related Experiment Videos

  • The method is designed to handle measurement error whose distribution depends on other covariates.
  • Performance was evaluated through simulations comparing it to regression calibration and multiple imputation.

    • Main Results:

    • The proposed SIMEX extension demonstrated superior performance in simulations, showing reduced bias and variability.
    • It effectively handled situations where measurement error distribution is dependent on other covariates.
    • The method was successfully applied to real-world data on HIV phylogenetic cluster size and sexual partnerships.

    Conclusions:

    • The extended SIMEX method offers a robust approach for bias reduction in the presence of complex measurement error.
    • This technique is valuable for epidemiological studies, such as analyzing HIV transmission dynamics, where validation data is scarce.
    • The study provides a practical tool for examining associations between variables affected by non-classical measurement error.