dear list.
i need help in interpretating results.
i designed a model with 2 factors (group (2 levels) x time (4 levels). in
my model i give in the first level results of my subjects, which are con-
files for 4 serial time points (T1 -T4)in relation to the baseline (T0).
in a roi, which is positive for the main factor of group, i want to plot
the activation over the time for the two groups.
now my question (i attached a jpg-file with 2 figures):
with help of the voi function in spm5 i can get data for each of my
subjects for this cluster and can create bars for each of my time point (T1-
T4). this gives me the impression of a negative course of activation (in
relation to the baseline)for my patients and positive for my controls (see
figure 1).
with help of the data extraction in marsbar i can calculate the % of signal
change of my 4 time points (T1-T4) in relation to the baseline. This gives
me the impression of positive changes in activation in both groups (figure
2).
now what is the difference between both approaches?
can i say that, in regard to figure 1 and the voi tool in spm, my patients
show a deactivation in this roi in relation to the baseline?
or is this stupid, because the con-files are just statistical images of the
beta of my glm-model and this gives no idea about the time course of
activation?
so would figure 2 be the answer?
hope i was able to pose my question in an understandable way.
thanks in advance
markus
|