Dear Chiara,
if A and B are your two clinical scores, you can add the interaction
term as an extra regressor, after mean-centering:
(A-mean(A)).*(B-mean(B))
Best regards,
Guillaume.
On 12/12/15 16:49, Chiara Pinardi wrote:
> Dear experts,
> I REALLY need some advises for my fMRI study based on a regression analysis with two scalar variables and 20 pathological subjects.
>
> From what I understand, SPM multiple regression tests the interaction between one covariate and the fMRI data correcting for the other one. I need to test also the interaction among the clinical scores on the fMRI data: how can I do? Can I set a F-test on the regression matrix to do that or do I need to set a new, special ANOVA test?
>
> Thanks in advance for any help!
> Chiara
>
--
Guillaume Flandin, PhD
Wellcome Trust Centre for Neuroimaging
University College London
12 Queen Square
London WC1N 3BG
|