Hi
possible, though essentially yu already know this now, given that the
3 different analyses gave you the same spatial components...
hth
christian
On 26 Mar 2008, at 20:45, Michelle Voss wrote:
> Hi all,
>
> I've run tensor-PICA on 3 block design tasks (same block order for
> everyone) separately. I get 5 replicable components of interest
> across across each task. My hypothesis is that the nature of these
> 5 components is independent of the tasks themselves. For one way to
> test this I was wondering if it's statistically appropriate to
> concatenate the filtered_func images for the three tasks, generating
> one 4D file for each subject which contains the three tasks in the
> same order for all subjects. Then, run tensor-PICA on these images.
>
> Would components estimated by this analysis appropriately represent
> common factors among the three tasks?
>
> best,
> Michelle
|