Hi Anderson,
I have a couple more questions:
1) You said in the previous message:
If age and sex don't change between the 3 measurements per subject, these don't need to be entered in design1, as the subject-specific EVs already account for that. For design2, you can add extra EVs for age and sex.
For the 3-way interaction of Age x Group x Emotion which has 3 designs, where should the sex covariate go - in design2, or in design3?
2) For the 3 way-interaction, should the continuous covariate of age be mean-centered in design3 or kept in raw form? Similarly, should sex covariate also be mean-centered?
Thanks!
Michelle
|