Jove
Visualize
Contact Us

Related Experiment Videos

On confidence intervals for within-subjects designs.

David C Blouin1, Arthur J Riopelle

  • 1Department of Experimental Statistics, Louisiana State University, Baton Rouge, LA 70803, USA. dblouin@lsu.edu

Psychological Methods
|January 6, 2006
PubMed
Summary

This study reveals that confidence intervals (CIs) and null hypothesis significance testing (NHST) conclusions can be inconsistent in within-subjects designs due to differing statistical models. Mixed models offer a consistent approach.

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

Effects of Volume, Pitch, and Phonation Type on Oscillation Initiation and Termination Phases Investigated With High-speed Videoendoscopy.

Journal of voice : official journal of the Voice Foundation·2016
Same author

Integrating flood depth and plant resistance with chlorantraniliprole seed treatments for management of rice water weevil, Lissorhoptrus oryzophilus (Coleoptera: Curculionidae).

Insect science·2014
Same author

Evaluation of alfalfa inter-seeding effect on bahiagrass baleage fermentation and lactating Holstein performance.

Journal of animal science and biotechnology·2014
Same author

Circumferential adjustment of ultrasound probe position to determine the optimal approach to the internal jugular vein: a noninvasive geometric study in adults.

Anesthesia and analgesia·2005
Same author

The difference between t and z and the difference it makes.

The Journal of general psychology·2004
Same author

Functional anatomy of the null hypothesis and of tests of it.

The Journal of general psychology·2003
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

Area of Science:

  • Statistics
  • Psychological Measurement
  • Biostatistics

Background:

  • Confidence intervals (CIs) are often proposed as alternatives to null hypothesis significance testing (NHST).
  • A key debate point is whether conclusions from CIs and NHST should align.
  • Inconsistencies arise in within-subjects designs depending on the underlying statistical models used.

Purpose of the Study:

  • To examine classes of CIs that yield inconsistent conclusions with NHST in within-subjects designs.
  • To investigate CIs that produce consistent conclusions with NHST.
  • To propose mixed model methodology for improved consistency.

Main Methods:

  • Analysis of fixed-effects models with subjects treated as fixed.
  • Comparison with mixed models where subjects are treated as random.

Related Experiment Videos

  • Exploration of generalizations to various within-subjects designs.

    • Main Results:

    • Identified specific classes of CIs leading to divergent conclusions from NHST in within-subjects designs.
    • Demonstrated that fixed-effects models (subjects fixed) contribute to this inconsistency.
    • Highlighted the consistency achieved through mixed models (subjects random).

    Conclusions:

    • The choice of statistical model (fixed vs. mixed effects) is critical for the consistency between CI and NHST conclusions.
    • Mixed model methodology provides a more coherent framework for analyzing within-subjects data.
    • Future debates on NHST should consider the implications of different modeling approaches.