Hi all. Although I don’t plan to use a mixed ANOVA, I am interested in this question due to 1) useful for power aproximation for other tests and 2) to better understand GPOWER bugs, which is what my peers use.

Question: Which power calculations are correct for a mixed ANOVA (2b x 2w)?

Results are identical for ANOVA 2b x 2b, but contrast almost by 50% for a mixed ANOVA (2b x 2w).

Thank you.

Brysbaert, Table 7

Sommet et al

This question might be a bit off topic for this forum, but here I go.

I think both are making different assumptions and are looking at different components for power. Both could be “right” under the conditions they are simulating.

If you’re savvy with R, I have some documentation on mixed ANOVAs for my Superpower R package.

1 Like

Thanks Aaron. Sorry, but I didn’t receive notification of your reply.

Yes, I am somewhat familiar with your package.

And yes, i also start to agree this is the problem.

Both could be “right” under the conditions they are simulating.