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

Estimation of the correlation between nutrient intake measures under restricted sampling.

C Y Wang1, G L Anderson, R L Prentice

  • 1Division of Public Health Sciences, Fred Hutchinson Cancer Research Center, Seattle, Washington 98109-1024, USA. cywang@fhcrc.org

Biometrics
|April 21, 2001
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

[Role of red nucleus in inhibiting nociceptive responses of rat spindle afferent].

Sheng li xue bao : [Acta physiologica Sinica]·2001
Same author

Suppression of tumor necrosis factor-mediated apoptosis by nuclear factor kappaB-independent bone morphogenetic protein/Smad signaling.

The Journal of biological chemistry·2001
Same author

Colonoscopic manifestations of primary colorectal lymphoma.

Endoscopy·2001
Same author

Steroid allergy: report of two cases.

Journal of microbiology, immunology, and infection = Wei mian yu gan ran za zhi·2001
Same author

A second deciduous molar impacted in right maxillary sinus: a long-term follow-up.

Chang Gung medical journal·2001
Same author

Metabolism of phenolic compounds during loquat fruit development.

Journal of agricultural and food chemistry·2001

This study addresses biased correlation estimates from food frequency questionnaires (FFQs) in dietary intervention studies. Likelihood-based and multiple imputation methods accurately adjust for restricted participant selection, improving dietary assessment reliability.

Area of Science:

  • Epidemiologic Research
  • Nutritional Science
  • Biostatistics

Background:

  • Food frequency questionnaires (FFQs) are standard for assessing dietary intake in research.
  • Evaluating FFQ reliability typically involves correlating FFQ data with food records (e.g., 4-day food records [4DFR]).
  • Dietary intervention studies may select participants based on baseline FFQ intake, creating a restricted population.

Purpose of the Study:

  • To address biased correlation coefficient estimation in dietary intervention studies due to restricted participant selection.
  • To present and evaluate statistical methods for handling incomplete dietary data arising from study design.
  • To improve the accuracy of dietary intake assessment in research settings.

Main Methods:

  • Application of likelihood-based methods to accommodate incomplete dietary data.

Related Experiment Videos

  • Utilizing multiple imputation (MI) techniques for data adjustment.
  • Conducting a simulation study to assess the performance of various estimators.
  • Analytical investigation of the robustness properties of maximum likelihood (ML) estimators.

    • Main Results:

    • Likelihood-based and multiple imputation methods effectively adjust for restricted populations in dietary studies.
    • Both MI and ML estimators show resilience to departures from normality assumptions.
    • The proposed methods provide less biased estimation of correlation coefficients and other parameters.

    Conclusions:

    • Statistical adjustments are crucial when analyzing dietary data from selected participants in intervention studies.
    • Likelihood-based and MI approaches offer robust solutions for incomplete dietary assessment data.
    • These methods enhance the reliability of dietary intake assessments in epidemiologic research, as demonstrated in the Women's Health Initiative study.