hi all,
i am trying to perform a conjunction across conditions A and B for a
single group of subjects, at the random effects level. i would be very
grateful if anyone could please point out if i am doing the correct
sequence of steps.
1) set up a basic model (one way anova, 2 groups).
2) choose the group 1 images ( which are condition A contrasts , con * .imgs).
Question : is this correct, or could i choose the SPMT_***.img s here?
3) similarly , choose group 2 images ( condition B)
4) choose non sphericity correction , replication over groups(2),
correlated repeated measures ( since contrast A and contrast B images
are from the same group of subjects)
5) estimate the design.
6) then specify two contrasts in the contrast manager as 1 0 and 0
1 ( for conditions A and B) . then select both the contrasts
using Ctrl and test for the ConjNull hypothesis using Tom Nichol's
spm2_conj plugin.
many thanks for any suggestion!
bhisma
----------------------------------------
Bhismadev Chakrabarti
Department of Psychiatry
University of Cambridge
Cambridge, UK