On Nov 9, 2009, at 12:22 PM, Dav Clark wrote:
> 2) or perhaps I should do the first part according to the
> instructions and then just select those 5 cope directories for each
> subject in the fixed effects gfeat directory. (this is what I did)
>
> It's not clear to me if there'd be any difference mathematically in
> the above - perhaps some correction for multiple comparisons in the
> latter?
I should add, however, that (2) seems to just give a meaningless grand-
mean of all your contrasts for a given subject. So it seems this only
works if you have a single contrast of interest.
I'm actually modeling 8 different conditions (actually 2 x 2 x 2) and
computing 8 contrasts of interest.
Thus, it seems that I need to in fact do something different than the
second part of the random effects documentation! Doing a fixed effect
analysis as suggested in the #MultiSessionMultiSubject section doesn't
leave me with separate copes I can select for each subject, so it
seems I need to do a fixed effect analysis on each subjects runs
separately, then grab the relevant cope across each of those subjects'
directories for a separate mixed-effects analysis per contrast.
Or am I missing something?
The particularly odd thing is that my fixed-effect analysis as per
#MultiSessionMultiSubject does actually present my results organized
in a useful way - with a page per contrast showing the fixed-effect
average across runs for each subject.
Cheers,
Dav
|