David B. Wilson, PhD, George Mason University
Version date: 2023.11.27
ANOVA: Two-way Analysis of Variance
Notes: This method is for a two-way ANOVA where both factors (independent variables) are between subjects. For example, one factor might be treatment, and the other might be the risk level of the participants as a categorical variable. One factor must represent the treatment contrast of interest and have only two conditions (treatment and control). The other factor may have any number of conditions (or groups, levels, etc.). The default method is to add the variance associated with the second factor into the denominator of the effect size (i.e., the pooled within groups standard deviation). This approach is sensible if the factor reflects a characteristic of the study participants, like gender or risk level. In contrast, if the second factor is experimental (i.e., a manipulation), then it may not be sensible to do so, in which case, change the default selection at the top of the page regarding factor variance.

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