Dear Christian Beckmann,
>> suppose I have three conditions EX1, EX2, and N and let's assume I
>> define two 1st-level contrasts as '1 0 0' and '-1 0 0'.
>>
> Fine, though if you denote by N the 'null condition' or baseline then
> note that all time series get de-meaned in FEAT prior to the analysis.
> You therefore would not model this explicitly.
Actually, 'N' stands rather for a control condition.
But ok, I guess that is exactly the crucial point for me: Could you give
me a hint where to read about the process of 'de-meaning'?
Thanks!
-Lucas
|