David B. Wilson, PhD, George Mason University
Version date: 2023.11.27
ANCOVA: Analysis of Variance with a Covariate
Notes: This method is for an ANCOVA model with a factor for the treatment contrast comparing two means and a covariate, such as a pre-test of the outcome variable. It is unsuitable for more complex models with multiple factors or covariates. This effect size from this method is not adjusted for the covariate. It uses the MS error and the correlation between the covariate and the dependent variable to recover the pooled within-groups standard deviation.

This work is licensed under a Creative Commons Attribution-Noncommercial 3.0 United States License.
