Dear Caroline,
In your case, use a two-sample t-test design and add the covariate with
option "Interactions" "With factor 1". As for centering options, have a
look at this document from J. Mumford (the last example):
http://mumford.fmripower.org/mean_centering/
Best regards,
Guillaume.
On 30/08/17 10:15, Caroline L wrote:
> Dear SPM users,
> I want to find out which part of the brain activation is more correlated with a behavioral skill in group 1 and in group 2 separately.
> My behavioral skill is a score from 0 to the infinite (gamma distribution) better is the subject lower is the score.
>
> I want to use a multiple regression analysis in SPM12 but I am not quite sure how to do this. I am confused about how to use some parameters :
> factorial design specification > Design : Multiple regression
> * Covariate > Centering or no centering? What is the difference between both and in which case should I use either ?
> * Intercept > Include Intercept or omit intercept? What is the difference between both and in which case should I use either ?
>
> Thank you in advance for your help
>
> Best regards
>
> Caroline
>
>
>
>
--
Guillaume Flandin, PhD
Wellcome Trust Centre for Neuroimaging
University College London
12 Queen Square
London WC1N 3BG
|