Hi Matthijs,
I suppose you consider an activation study. If so you do not need to write
all zeros, and consider [-1 1]. You multiply this contrast by the
parameters matrix beta, and get a simple t_a=(meanA_a-meanB_a)/SD(AB)
for a given voxel 'a' and two datasets A & B.
Check the reference http://www.mrc-cbu.cam.ac.uk/Imaging/spmstats.html
for a step-by-step and well written description.
Best regards,
FS
[[log in to unmask] écrivait/wrote :]
> Dear SPM community,
> Where can we find information on how t-maps are generated from a
> contrast such as -1 1 0 0 0 ? As we understand it, the contrast-
> manager uses beta-values, right? For one factor, the multiple
> correlation can be calculated and used to calculate the t-values for
> this factor. But how about two factors?
>
> Thanks in advance for any helpful info
>
> Matthijs Vink
--
[log in to unmask] [@hcuge.ch] - BIOMED. & RESEARCH Eng.
PET & Neuroscience research group - Div. Nucl. Med. /Geneva Univ. Hosp.
Prof: 0(+041)22.37.27268 || (HUG): 27268 || Perso: 0(+033)6.18.66.38.68
|