See inline response below.
On Mon, Aug 27, 2012 at 12:56 PM, Bianca <[log in to unmask]> wrote:
> Dear SPM'ers,
>
> I have setup a flexible factorial model with three factors:
>
> 1. subject (18 subjects per measurement) (indepency=yes variance=equal)
> 2. time (3 measurements) (indepency=no variance=equal)
> 3. condition (4 different contrasts) (indepency=no variance=equal)
>
> I specifief all scans and added the following matrix (54*4):
>
> 1 1 1 1
> 1 1 1 2
> 1 1 1 3
> 1 1 1 4
> 1 1 2 1
> ...
> ...
> 1 18 3 3
> 1 18 3 4
>
> For the results I am interested in:
>
> a. whether the activation pattern at measurement 1 is the same as the activation patterns at measurements 2 and 3 for a specific condition (T1C1=T2C1=T3C1)
>>> This cannot be answered as you can not conclude that two conditions are statistically the same. You can only conclude conditions are statistically different. The absence of conditions being different does not imply that the condition are the same though.
> b. which brain areas are active for a specific condition over time (positive effect of C1 for example)
>>> This is a between subject effect and thus cannot be tested in this model.
>
> Now I have ceated two main effects (Time and Condition) and an interaction effect (time*condition).
>
> The problem is that I can't figure out how to define the correct contrasts (e.g. contrasts weights) and I am not sure whether I used the correct main and interaction effects.
>
> Can someone please help me?
>
> Thanks!
>
> Bianca
|