You shouldn't expect to see any specific differences in a between-group
map based on eyeballing the within-group maps. That is, just because you
see a significant difference in left prefrontal in the AS kids, and you
don't see that in the controls, doesn't mean that when you compare them
to each other there will still be a difference between the two.
Ana Catarino wrote:
> Hello,
>
> I'm running an SPM analysis of 24 subjects divided in 2 groups: 12 controls
> and 12 autistic subjects. I've done the pre-processing and 1st level analysis
> for each subject but I'm having a bit of a problem running the 2nd level
> analysis.
>
> I ran a 2nd level analysis for each group ('One-sample t-test' including the
> images for control subjects, and 'one-sample t-test' using the images for the
> autistic subjects) and realised that there were large significant differences
> between the two groups for a given '1st-level' contrast (see images attached -
> AS_group.pdf and Control_group.pdf). However, when I try to compare the 2
> groups directly, using a '2-sample t-test', the results seems to lose
> significance and I can't see the differences I expected to see based on
> the 'one-sample t-test' results. I used 3 different contrasts in the '2-sample t-
> test':
>
> F-contrast [1 -1] --- (ASvsControl(1).pdf)
> T-contrast [1 -1] --- (ASvsControl(2).pdf)
> T-contrast [-1 1] --- (ASvsControl(3).pdf)
>
> But I can't get the results I expect with any of them, although they seem the
> logical contrasts to use when trying to check the differences in activation
> between two different groups of subjects (but perhaps I'm not thinking
> correctly).
>
> If anyone could give me any clue on how to solve this problem, or if someone
> could tell me what I'm doing wrong I would be very very thankful.
>
> Best wishes,
> Ana Catarino
>
|