0
$\begingroup$

Experimental design: 3x12 repeated measures.

I ran a mixed effects model that revealed an effect of "variable x" for which there are 12 levels. Post-hoc analyses based on this model (paired samples t-tests) revealed a significant difference between level 1 and 12 (p < 0.05).

I would like to compute the sample size necessary to find such an effect in a replication experiment (alpha = 0.05, power = 0.8) using G*power. As such, I need to enter an effect size.

Do I use the effect size associated with the omnibus test (F), i.e. the "effect" of x, or the effect size associated with the post-hoc (t) test, i.e. the difference between level 1 and 12?

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.