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

Resampling techniques in the analysis of non-binormal ROC data

D Mossman1

  • 1Division of Forensic Psychiatry, Wright State University School of Medicine, Dayton, OH 45401-0927, USA.

Medical Decision Making : an International Journal of the Society for Medical Decision Making
|October 1, 1995
PubMed
Summary

Resampling techniques like the bootstrap offer robust alternatives for analyzing receiver operating characteristic (ROC) data when standard binormal assumptions fail. These methods are particularly useful for hypothesis testing and confidence intervals in small datasets.

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

Malpractice and the psychiatrist : a primer for residents.

Academic psychiatry : the journal of the American Association of Directors of Psychiatric Residency Training and the Association for Academic Psychiatry·2014
Same author

The -149C>T SNP within the DeltaDNMT3B gene, is not associated with early disease onset in hereditary non-polyposis colorectal cancer.

Cancer letters·2008
Same author

Avoiding errors about 'margins of error'.

The British journal of psychiatry : the journal of mental science·2007
Same author

"A fool for a client": print portrayals of 49 pro se criminal defendants.

The journal of the American Academy of Psychiatry and the Law·2002
Same author

Intervals for posttest probabilities: a comparison of 5 methods.

Medical decision making : an international journal of the Society for Medical Decision Making·2002
Same author

The meaning of malingering data: further applications of Bayes' theorem.

Behavioral sciences & the law·2001

Area of Science:

  • Biostatistics
  • Medical Informatics
  • Statistical Modeling

Background:

  • Receiver Operating Characteristic (ROC) analysis is a key tool for evaluating diagnostic tests.
  • Common ROC analysis methods rely on binormal assumptions for latent frequency distributions.
  • These assumptions may not hold true for all datasets, necessitating alternative approaches.

Purpose of the Study:

  • To describe the application of resampling techniques for ROC data analysis when binormal assumptions are inappropriate.
  • To highlight the utility of bootstrap methods for confidence intervals with small sample sizes.
  • To advocate for resampling as a versatile and distribution-independent approach.

Main Methods:

  • Application of jackknife and bootstrap resampling techniques.

Related Experiment Videos

  • Hypothesis testing using distribution-independent methods.
  • Confidence interval estimation for ROC data.

    • Main Results:

    • Resampling techniques provide appropriate methods for ROC data that do not fit the binormal model.
    • The bootstrap method is particularly effective for small data samples.
    • Faster computers enhance the accessibility and convenience of resampling methods.

    Conclusions:

    • Resampling techniques are valuable alternatives to binormal assumptions in ROC analysis.
    • The bootstrap method offers a robust approach for confidence intervals, especially with limited data.
    • Increased computational power makes resampling a practical tool for modern data analysis.