I've never powered something like this. But assuming you have information about the design, you should be able to do it with the generic F test calculator in G*Power.

With that said, are you set on repeated measures ANOVA? For most cases like this, some type of mixed/multi-level model or similar is better. And I'm sure you'd have more success finding resources to help power it.

Thank you Joe Stanek for your reply.

I will try your suggestion, although, again, I don’t think the G*power program is suitable becauae I have two within factors with each having two levels and when I tried to use the program I didn’t find the suitable rubrics to each of the measures, meaning that in the within factors option the program assumes that there is only one measure with x levels unless I can enter in this rubric the value 4 which is the total of the levels in both measures which I‘m not sure it’s the right way to do it statisticall...

I just thought with the general design, you would be able to know the numerator and denominator degrees of freedom, even with extremely complex designs, and then be able to use the generic F test to give a general idea of a sample size/power.

Hello :)

I have a very similar problem currently - 2 within factors with 2 levels each in a repeated measures design. Looking to find the suitable sample size given the effect size, power and alpha level.

I have looked at the generic F-test option in G*power, but this does not output any suggestions for sample size. Any ideas how to remedy this? Thanks in advance!