If you are comparing two groups, then the 2-sample t-test will be the
proper model, although other designs would work, too.
On Mon, 1 May 2006, Ozlem Ozmen Okur wrote:
> I am also looking for an answer to the same question.
>
> Thanks
>
> Ozlem OZMEN OKUR
>
>
> ----- Original Message ----
> From: Niels Focke <[log in to unmask]>
> To: [log in to unmask]
> Sent: Thursday, April 27, 2006 10:56:30 AM
> Subject: [SPM] Which model to use for VBM with SPM5
>
>
> I know that this may be a very dummy question but I did not find any
> advice or documentation from other sources.
> There has been a very substantial change in the GUI for the statistical
> models in the new SPM5. We are interested in a very basal group
> comparisons using VBM.
> 3D T1-Images where segmented, modulated and normalized with the new
> segmentation tool of SPM5 (unified model) and smoothed with 12mm Gaussian
> kernel.
> Unfortunately the basic models GUI of SPM2 is gone (allowing for simple
> ANOVA with e.g. 2 groups) and has been replaced with the 2nd level
> analysis or basis models (PET) but there is no obvious difference
> between the later two.
> My basic question is:
> Which model should we use for a simple 2-group comparison e.g. with
> dementia patients?
> One sample t-test and defining the groups with a covariant vector (0 or 1
> for group assignments).
> Or 2-sample t-test with groups assigned to group1 and 2 in the GUI
> Or even factorial designs?
>
> I really would appreciate any advice on this issue or maybe a suggestion
> where to find a documentation. The original SPM documentation does not
> help us very much on this issue.
>
> Thanks.
>
>
--
Volkmar Glauche
-
Department of Neurology [log in to unmask]
Universitaetsklinikum Freiburg Phone 49(0)761-270-5331
Breisacher Str. 64 Fax 49(0)761-270-5416
79106 Freiburg
|