Hi,
Are you talking about "two regressors" at the first or second level?
I'm not quite following what's happening yet.....
Cheers.
On 18 Jan 2008, at 20:07, Jack Grinband wrote:
> Hi Steve,
>
>> b) You could take the first-level zfstats and do simplified tests
>> on them
>> yourself, after resampling into standard-space. For example,
>> sum(Z)/sqrt(n) gives simplified fixed-effects, whilst doing simple
>> OLS on
>> Z stats probably gives some kind of random-effects-related analysis
>> (though maybe don't quote me on that!)
>
> I tried the fixed-effects analysis you had suggested for the F-test
> and I'm
> getting some confusing results. I have two regressors and I want to
> compare the number of voxels significant at the group level (fixed
> effects)
> for either regressor (i.e. the union of the significant voxels of
> the two
> regressors) versus the number of voxels significant in the F-test.
> For any
> given threshold, the number of significant voxels in the union
> should be
> less than or equal to the number of significant voxels in the F-test.
>
> However, I'm getting the opposite result. There are many more
> voxels in the
> union than the F-test. So I'm wondering what is different between
> the fixed
> effects group analysis for the individual regressors and the F-
> test? That
> is, is sum(Z)/sqrt(n) exactly correct?
> many thanks!!!
>
> jack
>
---------------------------------------------------------------------------
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director, Oxford University FMRIB Centre
FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK
+44 (0) 1865 222726 (fax 222717)
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve
---------------------------------------------------------------------------
|