Hi Todd,
You just need to load the copeX.feat directories in your 3rd level
analysis, rather than the gfeat folder. Make sure you load them in a
consistent order (e.g. all cope1.feat, cope2.feat, ..., cope10.feat from
subject 1, then subject 2, and so on), so that you can easily set your
contrasts. To make it easier to load all 150 inputs, you can use the
paste command.
Cheers,
Stéphane
Stéphane Jacobs - Chercheur post-doctorant / Post-doctoral researcher
Espace et Action - Inserm U864
16 avenue du Doyen Lépine
69676 Bron Cedex, France
Téléphone / Phone: (+33) (0)4-72-91-34-33
Todd Thompson a écrit :
> Hi, all. I'm about to do something really inefficient, and I thought
> I'd ask if there were a better way before I started:
>
> I have 15 subjects, each with 3 experiment sessions. Each session has
> many (let's say 10) contrasts that I ultimately want to examine at the
> group level.
>
> I've run 15*3=45 first-level analyses, which gives me 45
> SubjXSessionY.feat directories, each directory containing all of the
> 10 contrasts for that session.
> I've run 15 fixed-effects second-level analyses, one for each subject,
> which gives me 15 SubjX.gfeat directories. The nice thing about this
> step is that it automatically included every contrast from the first
> level analyses, so I didn't need to run 10 different second-level
> analyses per subject.
>
> Now I'd like to run my group analyses, and I'd really like it to do
> all 10 contrasts at once, keeping the contrast names and orders so
> that I can use the html report to get an overview of how my data are
> looking. But my lower-level directories are the second-level gfeat
> directories now, and FEAT complains that they're empty of COPEs. So,
> as far as I can tell, it looks like I need to run 10 3rd level
> analyses, instead of just a single one. And then I'll end up with a
> bunch of gfeat directories each containing one COPE that's lost its
> ties to the lower-level COPE names/definitions. Is this correct? Is
> there a convenient way to do this? I keep thinking I'm overlooking an
> obvious button somewehre, since it seems like this is perhaps the
> *most* typical fMRI analysis stream, but I couldn't seem to find an
> answer. My apologies in advance if this is just a searching failure on
> my part.
>
> Thanks!
> Todd
>
>
>
|