Dear Expert
I have very different number of participants in these 4 groups:
female patient / female control / male patient / male control
But I still conduct fMRI FEAT analyses comparing male vs female, patient vs control.
I have 2 questions, could you please take a look at them? Or perhaps suggest some readings? Thanks!
*I guess FEAT "does not" do anything to adjust for this issue of "unequal number issue"?
So that normally, we just accept this "unequal number issue" as a limitation.
But can we say that there is always higher change of type 2 instead of type 1 error?
*If FEAT really does not do anything for this "unequal number issue"....is there anything we can do to improve the situation?
That's why I thought "assuming different variances between groups" is going to help...I thought by doing so, the model will become similar to the "real" situation, but that's not the case?
Thanks so much.
|