Dear FreeSurfers,
I have two discrete and binary factors (Dx: patients vs. controls; Sex: female vs. male), one continuous variable (Age), one nuisance factor (IQ). If there is a 3-way interaction of Dx x Sex x Age, regressing out IQ, is the following interpretation correct?
Red/yellow: In females (vs. males), the Age effect of cortical thickening is stronger in patients (vs. controls). Blue/cyan: In females (vs. males), the Age effect of cortical thinning is stronger in patients (vs. controls).
Note. by stronger, I mean greater magnitude without sign.
Thanks, Daniel