Hello all,
I was wondering if I could seek feedback regarding a design I am setting up in the GLM gui. I have been asked to set up a 1-way, 3-level b/n Ss ANOVA, controlling for gender and age. The principal effect of interest is the main effect of group; to my knowledge there are no specific predictions regarding the pattern of the group effect.
My first guess at how to do this (see also the attached files):
- For the EVs, adapt the 2x3 b/n Ss ANOVA example from the GLM wiki page, to model gender and group.
- In addition to the M.E. and interaction contrasts, add in two contrasts for the positive and negative age effects, using a centered age regressor. No new F-tests are added.
Does this sound reasonable? And, will it be feasible to obtain corrected p's for the ME of group with randomise (I've been assuming it is)?
Thanks in advance --
Karin
|