Hi
I tried several ways to do that. But as far as I know, the basic models
UI generates more than just a design matrix. If you ask for mean voxel
value as a covariates, it is added to the structure, ...
If it is a new subject for each new matrix, I don't know a efficient way
- unless you tear apart the .m to execute the correct part -. But if you
test a subject from a base of subjects versus the rest of the base,
there is a way. If it is what you need, tell it to me, I'll explain it
to you ;)
BTW, your tests seem one vs group. As I'm looking for pro and cons of
this method, what are you waiting from SPM and how will you interpret
your results ?
Sincerely,
Matthieu Brucher
Yves Roy wrote:
> Dear SPMers:
>
> I created a design matrix for one subject (and SPM created a SPM.mat file).
> My design is exactly the same for the others subjects in my study.
>
> I copied and pasted this SPM.mat file into the corresponding directory for
> another subject, thinking that the design being the same this piece of
> information could be reused, but when preparing the data ("Model
> specification & parameter estimation")SPM did not bring me to the next
> step, i.e. ask me to select the scans for this subject.
>
> Does that mean that SPM.mat (which does not contain any subject-specific
> info, provided design is the same) is not reusable (in that manner) ?
>
> If it is reusable, how should I proceed?
>
> Thanks
> Yves Roy
>
>
|