> is there a way to test A & ~B conjunction in spm? i would like to see
a
> brain region activate only for one but not the other group of
subjects. no
> straighfoward way is offered in the software i am using. any hint will
be
> appreciated.
>
I am not sure which software you are using, but here is a suggestion for
SPM. If I were in your situation, I would mask A with B, using the
exclusive mask. I would use p=0.05 uncorrected threshold in the
exclusive mask, just to be on the conservative side. In this case I
would be restricting my analysis on A in areas where the test on B were
not rejected (with p>0.05 uncorrected).
When you are doing this sort of analysis, you have to be very careful.
In general, statistical tests are designed to detect signals, not the
lack thereof. You have to understand that if you fail to reject the null
hypothesis for B, that doesn't mean there is no signal. Perhaps you
didn't have enough power to detect signals for B even if there were
some. So it's a tricky thing to do.
-Satoru
|