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

Sample size requirements for precise estimates of reliability, generalizability, and validity coefficients.

R A Charter1

  • 1Department of Veterans Affairs Medical Center, Long Beach, CA 90822, USA. CHARTER.RICHARD_A@LONG-BEACH.VA.GOV

Journal of Clinical and Experimental Neuropsychology
|November 7, 1999
PubMed
Summary

Determining adequate sample size (N) for reliability studies is crucial. A minimum of 400 subjects is recommended for precise reliability coefficients (r), with larger Ns for validity studies.

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

Item difficulty of the tactual performance test location score.

Perceptual and motor skills·2002
Same author

Discrepancy scores of reliabilities of the WAIS-III.

Psychological reports·2002
Same author

"Damn the precision, full speed ahead with the clinical interpretation.".

Journal of clinical and experimental neuropsychology·2002
Same author

Meaning of reliability in terms of correct and incorrect clinical decisions: the art of decision making is still alive.

Journal of clinical and experimental neuropsychology·2002
Same author

Tactual performance test trials: internal consistency reliability using the Gilmer-Feldt coefficient.

Perceptual and motor skills·2002
Same author

Formulas for combining summary statistics.

Psychological reports·2001

Area of Science:

  • Psychometrics
  • Statistical Methods
  • Educational Measurement

Background:

  • Reliability coefficients (r) are essential for assessing measurement consistency.
  • Previous studies often used insufficient sample sizes (N).
  • Confidence intervals (CI) provide a range for the true reliability coefficient.

Purpose of the Study:

  • To investigate the precision of reliability coefficients (r) based on sample size (N).
  • To determine recommended sample sizes for various reliability and validity studies.
  • To highlight the importance of adequate sample sizes for accurate psychometric evaluations.

Main Methods:

  • Analysis of confidence interval widths for different reliability coefficients (retest, alternate-form, split-half, alpha, intraclass, interrater, validity) as a function of sample size (N).

Related Experiment Videos

  • Survey of sample sizes used in published reliability studies.
  • Application of confidence intervals for obtained test scores.

    • Main Results:

    • The width of the confidence interval for reliability coefficients decreases with increasing sample size (N).
    • A minimum sample size of 400 subjects is recommended for reliability studies.
    • Validity studies may require substantially larger sample sizes than reliability studies.
    • A significant proportion (59%) of published reliability studies utilized sample sizes below 100.

    Conclusions:

    • Adequate sample size is critical for precise estimation of reliability coefficients.
    • A minimum of 400 subjects is a practical recommendation for reliability research.
    • Underpowered studies are common, potentially leading to imprecise reliability estimates.
    • Confidence intervals offer a practical approach to assessing the precision of reliability measures.