Hi - yes, have a look at the randomise manual; the permutation testing
corrects for multiple comparisons across space by building the null
distribution of the max (across space) of any given test statistic,
such as cluster size of voxelwise statistic height.
If you really see _nothing_ in the maxc image that suggests that
something maybe went wrong in the analysis - you would normally expect
to see something, even if it doesn't reach statistical significance
(i.e. maxc>0.95). Have a look at the raw _tstat image and check that
its histogram looks reasonable.
Cheers.
On 14 Dec 2007, at 14:42, Zhang, Xiaochu (NIH/NIDA) [F] wrote:
> Hi FSL experts,
> With the help from you, I almost finished my data via TBSS. Now I
> got two kinds of data: “tbss_tstat” and “tbss_maxc_tstat”.
> From the manual, I knew the “tbss_maxc_tstat” is the cluster p-value
> images. However, there are no details of this multiple comparison.
> I am a little confused. Why I used very low threshold (e.g. 0),
> there are still not very many activated clusters.
> Could you please do me a favor and tell me how tbss do multiple
> comparisons across space?
> Thank you very much!
> Xiaochu Zhang Ph.D
>
> Visiting Research Fellow
>
> NIH/NIDA-IRP
>
> 5500 Nathan Shock Drive
>
> Baltimore MD 21224
>
>
>
> Tel: (410) - 550 - 1440 ext. 434
>
>
---------------------------------------------------------------------------
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
---------------------------------------------------------------------------
|