Hi
this amounts to a voxel-wise one sample t-test on the spatial IC maps,
yes. Across space you'd then run something like the mm approach
hth
christian
On 12 Apr 2008, at 14:48, Xi-Nian Zuo wrote:
> Thanks,Christian. I understand your suggestion on the group-level
> analysis
> as one alternative of the one-sample t-test. Am I right? Could I
> perform
> one-sample t-test on these spatially independent components? Maybe
> cant,
> because the non-gaussian assumption on the source random variables.
>
> Best,
> xinian
|