Hi - seems like something's wrong here - increasing tfce_tstat should
definitely lead to increasing tfce_p_tstat (it does for me!). Are you
definitely comparing images from the same run of randomise and the
same contrast?
Cheers.
On 20 Oct 2008, at 19:22, Ashley DeMarco wrote:
> Gwenaelle,
>
> We have been thresholding the p-tstat maps from .95 to 1 to observe
> the
> significant voxels, since 1-p is the convention used to generate the
> p-tstat
> maps.
>
> I think I miss stated what we are seeing on our maps in my previous
> post.
>
> As the values on the tfce p-tstat map increase, the tfce tstat map
> values
> decrease. Example: for one of our voxels our tfce p-tstat value is .
> 952 (p
> of 0.048) and our tstat map value is 16694, then on another voxel,
> our tfce
> p-tstat is .99 (p=.01) and our tfce tstat is 12686.
>
> (We aren't using corrp at this time because we have such a small n)
>
> Most of the other cases that we have looked at when we have run the
> randomise tool as mentioned had the proper relationship between the
> p-tstat
> and the tstat.
>
> So our 2 concerns are:
> 1. How can we get out tstat values down so they are appropriate for
> the p
> values we are getting
> 2. Figuring out why we got p-values and t-stats that had the opposite
> relationship on this run
>
> Thanks for all your help,
> Ashley
>
---------------------------------------------------------------------------
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
---------------------------------------------------------------------------
|