Effect Size Calculator

David B. Wilson, PhD, George Mason University
Version date: 2023.11.27

ANOVA: Two-way Analysis of Variance

Select one:
Ignore non-treatment factor variance
Add non-treatment factor variance to the pooled standard deviation

F-value, treatment factor

F-value, other factor

Degrees-of-freedom, other factor

F-value, interaction

MS-error

Mean

N

Treatment

Control

Cohen’s d =

Hedges’ g =

95% C.I. =

95% C.I. =

v =

v =

se =

se =

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.