I would like to compare multiple groups (3) of subjects who have each each undergone a fMRI session before and after an intervention. So I have a "PRE" and "POST" data set for each subject and want to compare the influence of the intervention between the different subject groups.
I guess this reduces to finding the "difference of a difference" since I am looking for inter-group differences of the PRE-POST difference. Could you please advice me how to set this model upin FEAT.
Is it a 3x2 ANOVA with one of the levels non-independent? If so, what contrast do I use for the influence of the intervention?
Are there any alternative ways of modelling this? Can I, for example, do this in randomise?
Best wishes
Geoff
***************
Medical Physics
UCL
UK
|