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

A simple randomisation procedure for validating discriminant analysis: a methodological note.

D G Wastell

    Biological Psychology
    |April 1, 1987
    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

    Designing for the home: a comparative study of support aids for central heating systems.

    Applied ergonomics·2008
    Same author

    Feedback quality and environmentally friendly use of domestic central heating systems.

    Ergonomics·2007
    Same author

    Multiple-task performance on a computer-simulated life support system during a space mission simulation.

    Acta astronautica·2001
    Same author

    A conceptual framework for designing micro-worlds for complex work domains: a case study of the Cabin Air Management System.

    Computers in human behavior·2001
    Same author

    Effects of training on short- and long-term skill retention in a complex multiple-task environment.

    Ergonomics·2001
    Same author

    Effective multidisciplinary communication in healthcare: cooperative documentation systems.

    Methods of information in medicine·2000
    Same journal

    Effects of transcranial electric stimulation on attentional functions in healthy adults: A meta-analysis.

    Biological psychology·2026
    Same journal

    Afterword to the Biological Psychology Special Issue:Contributions of the Vagus to Psychological Functioning and Health:Reflections upon genuine relations between vagus-nerve functioning and psychological processes.

    Biological psychology·2026
    Same journal

    Characterizing Individual Differences in SCR Responsivity: A Hybrid Mixture and Single-Trial Modeling Framework.

    Biological psychology·2026
    Same journal

    Your Heart Beats Next To Mine: Daily Physiological Synchrony Among Black And White ADRD Caregiver-Care Recipient Dyads.

    Biological psychology·2026
    Same journal

    Stimulus arousal level and savoring:an electrocortical investigation.

    Biological psychology·2026
    Same journal

    Neural response to reward and pathological personality dimensions.

    Biological psychology·2026
    See all related articles

    Discriminant analysis (DA) validation is complicated by its focus on exaggerating group differences. A new randomisation test provides accurate validation for stepwise discriminant analysis (SWDA), outperforming traditional methods.

    Area of Science:

    • Statistics
    • Machine Learning
    • Data Science

    Background:

    • Discriminant analysis (DA) aims to optimize classification by maximizing between-group differences.
    • This inherent bias in DA complicates the validation of its predictive accuracy.
    • Traditional validation methods like jack-knifing are unsuitable for stepwise DA (SWDA).

    Purpose of the Study:

    • To address the validation challenges in discriminant analysis, particularly when stepwise selection is used.
    • To introduce and evaluate a simple randomisation test for validating SWDA.
    • To discuss the advantages of randomisation tests over conventional significance tests.

    Main Methods:

    • Development and application of a simple randomisation test for validating SWDA.
    • Comparison of randomisation tests with orthodox significance tests.

    Related Experiment Videos

  • Review of non-parametric methods for estimating error rates of prediction rules.

    • Main Results:

    • The presented randomisation test provides correct decisions for SWDA.
    • Randomisation tests demonstrate general superiority over orthodox significance tests for validation.
    • Non-parametric methods offer viable approaches for estimating prediction rule error rates.

    Conclusions:

    • A novel randomisation test effectively validates SWDA, overcoming limitations of existing methods.
    • Randomisation testing is a superior approach for validating classification models with stepwise feature selection.
    • Further exploration of non-parametric error rate estimation is warranted for robust prediction rule assessment.