Hi everybody.
I performed a multiple regression second level analysis on my fMRI
data. I have 16 (1 x subject) scans, and 3 "covariates" (that indeed
are my variable of interest), with 16 values each (1 x subject).
So the final design matrix has 4 columns, one for the scan (blank),
and 3 for the covariates (white and grey).
Now the fact is that I have no idea how to set the contrasts to test
my hypothesis.
I mean: if I set 1 for the "scans' columns", 0 for the second and the
third column and -1 for the fourth, am I testing if there is a
negative correlation between BOLD signal and the fourth variable?
F.
--
Federico Nemmi
Dipartimento di Psicologia,
Università La Sapienza, Roma
|