Hi Geon-Ho
> Dear SPMer,
>
> I run three tasks, baseline(BA), activation1(AC1), and
> activation2(AC2) for each subject and have total 15 subjects. I tested
> BA<AC1, BA<AC2, and AC1<AC2 for each subject and automatically created
> in con_1.img, con_2.img, and con_3.img for each contrast and each
> subject. I would like to know a group effect of each contrast that was
> BA<AC1 or BA<AC2 over 15 subjects. For the 2nd lebel test, could
> anyone explain how to test using SPM2 to see the group effect?
>
> For each subject,
> BA AC1 AC2 con
> BA<AC1 -1 1 0 con_1
> BA<AC2 -1 0 1 con_2
> AC1<AC2 0 -1 1 con_3
You should find this information on the SPM website (at least in the
user guide) or you can find it on my pages http://membres.lycos.fr/wamcyril/
In general select basic model to perform a 2nd level analysis. In your
case simply select one sample t-test - do that 3 times for each contrast
you are interest in. You should also try a paired t-test with AC1 and
AC2 to oppose them (AC1>AC2 and AC1<AC2) .. and possibly compare the
result with the one sample t-test from the con_3 images
hope this helps
best
cyril
|