Dear Bas,
if you don't need the fMRI specific stuff like AR(1) error covariance
structure and high-pass filtering, then you should be fine using a
simple t-Test. Also, you could try to model your constant term
explicitly as regressor.
Volkmar
Am Donnerstag, den 17.01.2008, 13:17 +0100 schrieb Neggers, S.F.W.:
> Dear list,
>
> for some obscure reason I want to estimate a GLM only containing only the constant term using SPM5. When I remeber well, in SPM2 I could do this by simply specifying zero conditions. In the SPM5 gui or batch mode, things work differently. To create a model only containing the baseline, I tried leaving the 'Condition' field under Data & Design in the model specification menu empty, and specified scans. I assumed SPM5 would now simply add the constant term to my model as it always does, and things would be groovey. However, when estimating the model, the following happened:
>
> Initialising parameters : ...computing??? Error using ==> mtimes
> Inner matrix dimensions must agree.
>
> Error in ==> spm_FcUtil>sf_X1o at 743
> b = spm_sp('ox',sX)*Fc.X1o.ukX1o;
>
> Error in ==> spm_FcUtil at 301
> varargout = {sf_X1o(Fc,sX)};
>
> etc etc...
>
> So apparently SPM expects onsets etc in a Condition field, no matter what.
> Does anyone now how to specify the simplest of all designs, namely one with only a constant?
>
> Any advice will be grealy appreciated.
>
> Regards,
>
> Bas
>
>
> --------------------------------------------------
> Dr. S.F.W. Neggers
> Division of Brain Research
> Rudolf Magnus Institute for Neuroscience
> Utrecht University Medical Center
>
> Visiting : Heidelberglaan 100, 3584 CX Utrecht
> Room B.01.1.03
> Mail : Huispost B01.206, P.O. Box
> 3508 GA Utrecht, the Netherlands
> Tel : +31 (0)30 2509609
> Fax : +31 (0)30 2505443
> E-mail : [log in to unmask]
> Web : http://www.fmri.nl/people/bas.html
> --------------------------------------------------
--
Volkmar Glauche
Freiburg Brain Imaging
http://fbi.uniklinik-freiburg.de/
Phone +49(0)761 270-5331
Fax +49(0)761 270-5416
|