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

Maximizing accuracy and precision using individual and grouped exposure assessments

N S Seixas1, L Sheppard

  • 1Department of Environmental Health, University of Washington, Seattle 98195, USA.

Scandinavian Journal of Work, Environment & Health
|April 1, 1996
PubMed
Summary

Random errors in exposure data can weaken exposure-response relationships. Combining individual and group exposure estimates helps control bias and imprecision, improving accuracy in health outcome studies.

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

Optimizing Air Pollution Exposure Assessment with Application to Cognitive Function.

Research report (Health Effects Institute)·2025
Same author

Impact of preferential sampling on exposure prediction and health effect inference in the context of air pollution epidemiology.

Environmetrics·2018
Same author

The use of cost-utility analysis for the evaluation of caries prevention: an exploratory case study of two community-based public health interventions in a high-risk population in the UK.

Community dental health·2018
Same author

Public perception of physiotherapy: implications for marketing.

The Australian journal of physiotherapy·2014
Same author

Parent satisfaction with the Minimal Motor Dysfunction Unit: a survey.

The Australian journal of physiotherapy·2014
Same author

Exposure assessment and public health.

New solutions : a journal of environmental and occupational health policy : NS·2012

Area of Science:

  • Environmental Health
  • Epidemiology
  • Biostatistics

Background:

  • Exposure assessment is crucial for understanding health outcomes.
  • Random errors in exposure data can bias exposure-response relationships.
  • Different exposure assessment methods have varying impacts on these relationships.

Purpose of the Study:

  • To evaluate the impact of random errors in exposure data on exposure-response relationships.
  • To compare individual, grouped, and combined exposure assessment methods.
  • To identify methods that minimize bias and imprecision.

Main Methods:

  • Utilized Monte Carlo simulations with 100 subjects per study, divided into four exposure groups.
  • Generated individual exposure data with assumed inter- and intraindividual variances (lognormal distribution).

Related Experiment Videos

  • Calculated individual mean, group mean, and hybrid (James-Stein shrinkage) exposure estimates for regression analysis against health outcomes.

    • Main Results:

    • Individual exposure estimates showed significant attenuation of the exposure-response relationship, especially with high within-subject variability.
    • Group mean estimates effectively controlled attenuation but reduced precision.
    • The hybrid estimator demonstrated the ability to simultaneously control both bias and imprecision.

    Conclusions:

    • Individual exposure means can lead to attenuated exposure-response relationships.
    • Grouped estimates mitigate bias but may decrease precision.
    • Combined individual and group estimates offer a way to control both bias and imprecision, though further research on various error structures and methods is needed.