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

Covariate imbalance and conditional size: dependence on model-based adjustments

S E Maxwell1

  • 1Department of Psychology, University of Notre Dame, IN 46556.

Statistics in Medicine
|January 30, 1993
PubMed
Summary

Analysis of covariance (ANCOVA) helps adjust for imbalances in randomized studies. However, accurate ANCOVA models are crucial for reliable results, and stratified assignment can improve this reliability.

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

Optical and mechanical design of a telescope for lunar spectral irradiance measurements from a high-altitude aircraft.

The Review of scientific instruments·2020
Same author

Precise methane absorption measurements in the 1.64 μm spectral region for the MERLIN mission.

Journal of geophysical research. Atmospheres : JGR·2016
Same author

Multitrait-Multimethod Comparisons Across Populations: A Confirmatory Factor Analytic Approach.

Multivariate behavioral research·2016
Same author

On Using Analysis Of Covariance In Repeated Measures Designs.

Multivariate behavioral research·2016
Same author

Multiheterodyne spectroscopy with optical frequency combs generated from a continuous-wave laser.

Optics letters·2014
Same author

Toxicity and exposure of an adenovirus containing human interferon alpha-2b following intracystic administration in cynomolgus monkeys.

Gene therapy·2011

Area of Science:

  • Statistics
  • Biostatistics
  • Experimental Design

Background:

  • Analysis of covariance (ANCOVA) is commonly used in randomized designs to address covariate imbalance.
  • ANCOVA aims to control the conditional size of hypothesis tests.

Purpose of the Study:

  • To evaluate the reliance of ANCOVA's effectiveness on model accuracy.
  • To explore alternative methods for improving ANCOVA's performance.

Main Methods:

  • The study examines the impact of ANCOVA model accuracy on hypothesis testing.
  • Residual analysis and subject matter expertise are highlighted as key for accurate ANCOVA models.
  • Stratified assignment procedures are investigated as a means to reduce model dependence.

Main Results:

Related Experiment Videos

  • ANCOVA's ability to control conditional test size is highly dependent on the accuracy of the chosen statistical model.
  • Thorough examination of residuals and incorporation of domain knowledge are essential for building accurate ANCOVA models.
  • Stratified assignment procedures were found to decrease the sensitivity of conditional size control to specific ANCOVA model choices.

Conclusions:

  • Accurate ANCOVA models are critical for valid statistical inference in randomized studies.
  • Stratified assignment offers a robust approach to mitigate issues arising from ANCOVA model misspecification.
  • Researchers should prioritize model diagnostics and consider stratified randomization for enhanced control in hypothesis testing.