Ask a question
IBM TechXchange 2025 conference is accepting
Session proposals through April 11
I'm working on Univariate model with three variables: dependent variable, independent variable and a covariate. The model requires estimates for the effect size, preferably in the form of Cohen's d.
Checking SPSS support, it is possible to calculate the "d" coefficient from the partial squared eta calculated in the ANCOVA regression. However, it assumes the number of observations for the two groups (dependent and independent) are equal which is not my case.
Does anybody have any suggestions for the calculation of Cohen's d?
This seems to be a controversial topic. See
https://www.researchgate.net/post/Is_Cohens_d_an_appropiate_effect_size_for_a_two_group_comparison_with_ANCOVA
I'm new here and I don't know how to reply to Jon Peck's answer to my question. I am really grateful for the link and it certainly provides the information I needed. Thank you Jon