2009/10/28 David Soto
<[log in to unmask]>
Hello-
This is a question of a FSL beginner I am
I want to demean the Reaction Time data to be included as a modulator of
relevant EVs in the 3 Custom format starting at the first leve
I have 12 subjects and each has 4 runs
A) I understand that demeaning equals mean RT - individual RT
(or does it matter if one does individual RT - mean RT)
You want to demean just within each model. If you are demeaning a first level regressor with a variable amplitude (third column) and constant durations, then simply take the mean of the amplitude and subtract this from the amplitude of each response. If you did it the other way round (mean - resp), it would simply flip the sign of the fits of that regressor, affecting which contrasts (+ve or -ve) would show effects.
Eugene