Hi FSLers,
Im doing a three group comparison using a one level three factors model
(groups are composed by 15, 37 and 44 subjects respectively). I used the one
factor four levels example as my guide to create my model. In the feat
manual it says that to compare a level with another you can have one EV per
level. However, if you want to ask the ANOVA question - where is there any
treatment effect which is what im interested in you need to do:
EV1 fits condition C (it is the only nonzero EV during condition D). EV2
fits A relative to this, i.e. represents A-C (see below for explanation).
The F-test then tests for any deviation - ie any difference between the
levels, and corresponds exactly to the standard ANOVA test.
If, as well as the ANOVA test, you wanted to interpret individual contrasts:
If we define m,a,b,c as the 4 PE values, then
A=a+m, B=b+m, C=m.
Thus the first PE, m, is level C, the second is a=A-m=A-C, etc.
To get the mean: mean = (A+B+C)/3 = m+(a+b)/3 = contrast [ 1 1/3 1/3 ] This
is mean activation during the task right?
To get A-mean: A-mean = (2a-b)/3 = contrast [ 0 2/3 1/3 ] Is this right?
To get C-mean: C-mean = (-a-b)/3 = contrast [ 0 -1/3 1/3 ] Is this right?
I wondered too what contrast would be B-mean???
I also included demeaned age, sex and handedness as covariates and I am not
quite sure if this is right. Also when I used randomise executing the same
demeaned covariates in the design.con and .mat files I obtained very
different result from feat. Why is that? Im very sorry for this long thread
but Im quite stuck here.
Thank you for your immense help.
Best wishes.
--
Andres Roman
|