Hello all SPMers,
I have a data set consisting of patients and healthy controls. I
have performed both one-sample and two sample second level analyses
to examine the effects of my tasks on the groups independently and to
compare one group to the other.
My first question is that if I do one group alone as a one sample
t-test I get more robust and slightly different results than if I do
that group alone as part of the two sample t-test analysis (i.e. 1 0
1 or 0 1 1). Are both of these valid ways to do the analysis? I
assume the single group alone is more robust because there is less
variance.
I am also interested in seeing what the two groups have in COMMON.
I tried doing the conjunction of 1 0 1 (patients) and 0 1 1
(controls) in the two sample t-test. Is this right? Is there another
way to get at this question?
thanks very much!
alex
|