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 cumulative odds ratios.

Stephen R Cole1, Paul D Allison, Cande V Ananth

  • 1Department of Epidemiology, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD 21205, USA. scole@jhsph.edu

Annals of Epidemiology
|March 24, 2004
PubMed
Summary

This study introduces a new statistical method for estimating ordered odds ratios, relaxing the assumption of homogenous exposure effects. The approach allows for more flexible and realistic assessment of etiologic effects across response levels.

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

Contributions of preeclampsia to preterm delivery: a hypothetical interventional cohort study using incremental propensity scores.

International journal of epidemiology·2026
Same author

Ischemic Placental Disease and Hospitalization for Neurological and Psychiatric Disorders.

American journal of obstetrics and gynecology·2026
Same author

Infertility treatment is associated with peripartum hysterectomy: a population-based cross-sectional study.

Fertility and sterility·2026
Same author

Gestational diabetes and subsequent risk of type 2 diabetes and metabolic dysfunction-associated steatotic liver disease in a commercially insured U.S. Pregnancy cohort.

Diabetes research and clinical practice·2026
Same author

Cardiovascular Disease in Singleton Offspring Born of Pregnancies Complicated by Placental Abruption: A Population-Based Retrospective Cohort Study.

Journal of the American Heart Association·2026
Same author

Contribution of hypertension to recent changes in cardiovascular disease mortality: a population-based study in the United States.

Journal of hypertension·2026

Area of Science:

  • Biostatistics
  • Epidemiology
  • Statistical Modeling

Background:

  • Standard estimation of ordered odds ratios assumes homogenous etiologic effects across response thresholds.
  • This assumption is often unrealistic in real-world epidemiological studies.
  • A need exists for methods that relax this constraint for more accurate exposure effect assessment.

Purpose of the Study:

  • To present a novel statistical method for estimating ordered odds ratios.
  • To relax the constraint of homogenous etiologic effects of exposure across response thresholds.
  • To provide a more flexible approach for assessing exposure effects in ordered response data.

Main Methods:

  • The proposed method expands observed data into "person-thresholds."
  • Each subject is assigned a separate record for each response threshold.

Related Experiment Videos

  • Binary logistic regression is applied to estimate generalized cumulative odds ratios for exposures.

    • Main Results:

    • The method demonstrated increased flexibility in assessing etiologic effects of exposures in two examples.
    • Monte Carlo simulations confirmed that estimated cumulative odds ratios are unbiased.
    • Generalized estimating equations ensured proper confidence interval coverage in simulations.

    Conclusions:

    • The developed method offers straightforward estimation of ordered odds ratios.
    • It enables the etiologic effects of exposure to vary across different levels of an ordered response.
    • This provides a more nuanced understanding of exposure-response relationships.