Dear SPMers,
My experimental design has two between-subject factors with two levels
for each. So factor 1 = genotype (transgenic or wildtype) and factor 2 =
treatment (drug or water). The subjects were then scanned while
performing a simple on/off task. I have used the 'full factorial' design
with
the groups entered in the following order: (wt; water) (wt; drug)
(transgenic; water) (transgenic; drug)
To look for a main effect of genotype I use F contrast [1, 1, -1, -1]
to look for a main effect of treatment I use F contrast [1, -1, 1, -1]
and to look at the treatment*genotype interaction I use [1, -1, -1, 1]
So my first quesiton is, is this correct?
When looking at the SPM of the interaction I get a significant cluster.
To investigate the direction of the effects for the different groups
within this cluster is it ok to then extract the parameter estimates
(eg using marsbar) and analyse this outwith SPM.
Thanks for any help I really appreciate any advice you can give me,
Anjie
--
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.
|