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

Prior specification in Bayesian statistics: three cautionary tales.

Stefan Van Dongen1

  • 1Group of Evolutionary Biology, Department of Biology, University of Antwerp, Groenenborgerlaan 171, B-2020 Antwerp, Belgium. stefan.vandongen@ua.ac.be

Journal of Theoretical Biology
|March 21, 2006
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

Reply to Letter to the Editor on "The Added Value of Home-Uroflowmetry: Analyzing an Individual's VV-Qmax Curve and Inflection Point".

Urology·2025
Same author

Reply to Letter to the Editor on "The Added Value of Home-Uroflowmetry: Analyzing an Individuals' VV-Qmax Curve and Inflection Point".

Urology·2025
Same author

Reply to Editorial Comment on "The Added Value of Home-Uroflowmetry: Analyzing an Individuals' VV-Qmax Curve and Inflection Point".

Urology·2025
Same author

The Added Value of Home-Uroflowmetry: Analyzing an Individuals' VV-Qmax Curve and Inflection Point.

Urology·2025
Same author

No evidence for an association between facial fluctuating asymmetry and vocal attractiveness in men or women.

Evolutionary human sciences·2023
Same author

The contribution of teacher, parental and peer support in self-reported school and general well-being among ethnic-cultural minority and majority youth.

Frontiers in psychology·2023

Bayesian inference combines prior beliefs with data. Careful specification of prior distributions is crucial, as even weak priors can significantly bias results, especially in small samples.

Area of Science:

  • Statistics
  • Computational Biology
  • Biostatistics

Background:

  • Bayesian inference integrates prior knowledge with observed data.
  • Prior distributions represent beliefs before an experiment.
  • Bayesian analysis is increasingly popular due to computational advances.

Purpose of the Study:

  • To emphasize the critical importance of carefully specifying prior distributions in Bayesian analysis.
  • To demonstrate how inappropriate or uninformative priors can introduce substantial bias.
  • To advocate for data sharing to enable reproducible Bayesian analyses.

Main Methods:

  • Illustrative examples are used to demonstrate the impact of prior specification.
  • The study focuses on the interplay between prior distributions, likelihood functions, and posterior distributions.

Related Experiment Videos

  • The concept of 'complete ignorance' in prior specification is critically examined.

    • Main Results:

    • Inappropriate use of 'flat' or uninformative priors can lead to significant bias in posterior distributions.
    • Even weak prior information, when not carefully chosen, can unduly influence results, particularly in small sample sizes.
    • The study highlights that true 'absence of prior information' is often an illusion.

    Conclusions:

    • Prior distribution specification in Bayesian analysis requires careful consideration and biological context.
    • Researchers should exercise caution with seemingly uninformative priors, as they can still impart bias.
    • Promoting open data sharing is essential for transparency and allowing researchers to apply their own priors.