Hello,
The comparison that you want could be carried out by taking the individual
con images from the A minus rest and from the B minus rest through to a
paired t test at the group level.
However, it's worth noting that [(A minus Rest) minus (B minus Rest)] is
identical to (A minus B)
Best wishes
Paul Fletcher
On Oct 29 2006, Shaquanda Jones wrote:
>Hi SPMers,
>
> I'm trying to do the correct test while using rest as a control
> condition. In the experiment I have two main conditions of interest that
> I'll call A and B. There is also a rest condition where participants were
> looking at cross hairs. I would like to do the test of A minus rest
> compared to B minus rest. I'm a bit unsure about the correct way to
> specify this within SPM 2 however. Is this correct?
>
> 1) Start with basic models using a one sample t-test with A minus rest
> from individual's con images (using the contrast code "1, 0") 2) and a
> separate t-test for B minus rest also from individual's con images (also
> using the contrast code "1, 0") 3) This will give me two additional con
> images one representing A minus rest and the other representing B minus
> rest 4) Would I then do an additional one sample t-test comparing the two
> con images created in step 3 (using the contrast code "1, 0")?
>
>Thanks for your time,
>Sincerely,
>Shaqunada
>
>
>
|