> I have an fMRI study to analyze covarying for age. We have 2 groups,
> control(n=20) and patient(n=14) each with 8 conditions. These conditions
> are modeled in the 1st stage fixed effects level for each subject
> Is it reasonable to model all the condition and subjects in one anCova
> model in the second stage?
sounds reasonable to me
> Would I state 2 groups, 1st has 20*8=160 con.img files the 2nd group
> having 14*8=112 con.img files ?
> Or would I state 16 groups?: condition1group1 condition2group1
> condition3group1.....condition1group2 condition2group2.....
> Is it more correct to compare each condition within a 2 group design
> seperately, so I would run 8 seperate ancova analyses.
Well depends what you intend to do. But say you want to look at
everything, then 1 big AnCova should be fine - in SPM5 it is fairly easy
setting up a full factorial + 1 covariate (see also that you can add
more covariate by computing the interaction with factors). This
regressor should explain BOLD variations which follow age variations
across the two groups thus leaving the rest of your model 'cleaned' for
this effect.In SPM2 (which seems to be what you are using) I would
specify 16 groups so that you can compare each patient vs control for
each condition and their interactions, having the effect of the other
conditions and age controlled. (but the variance will not be partitioned
as good as it should be).
Hope this helps (and correct)
> Thanks for your time and attention.
Dr Cyril Pernet,
fMRI Lead Researcher SINAPSE
SFC Brain Imaging Research Center
Department of Clinical Neurosciences
University of Edinburgh
Western General Hospital
[log in to unmask]