Dear Susan,
|How do I accomplish this in SPM99? There is a default
|for an upper tail F probability and then when performing
|the stats analysis there is an option to correct the
|threshold and another option to change the p-value.
The upper tail F probability in SPM99 is a global threshold that restricts
the voxels for which data is saved (for further inspection/plotting
etc) during the analysis. I would be inclined to leave this as the default
value unless you have a specific reason to alter it.
After you have performed the analysis, the Results section allows
interactive threshold setting for inspection of the results. If you are
using voxel level inference, it would be conventional use a
spatial extent threshold of 0 (voxels; the inference is being made on
height, not spatial extents and there is no a priori reason to exclude
single voxels) and a height threshold corresponding to P<0.05
(corrected). This is quite a severe correction and it is usual that you
will have anatomically specified areas where you have a prior
hypothesis, and a lesser correction is appropriate. For voxels like this,
it would be appropriate to lower the threshold to P<0.001
(uncorrected) or use a small volume correction. Note that this is a
general, non-specific answer and if you have a specific experimental
question (or are using the cluster level of analysis, or small volume
correction) then they may well be different!
|Also, is there and option to view z-maps as an
|alternative to the F contrasts? There is an option for a
|t-statistic however I do not know how to define the
|contrasts.
Correct - although I believe that SPM99 generates SPM{T} maps
directly. The way in which you define contrasts depends on the
experimental question - you need to provide a few more details of
your experimental design and the question you're interested in
answering! You might also find Matthew Brett's web tutorial on contrasts
helpful; it's at http://www.mrc-cbu.cam.ac.uk/Imaging/spmstats.html
best wishes,
Geraint
-----------------------------------------------------------------
Dr. Geraint Rees
Wellcome Advanced Fellow Lecturer
California Institute of Technology Institute of Neurology
Division of Biology 139-74 University College London
Pasadena 12 Queen Square
California 91125 London WC1N 3BG
voice (626) 395-2880 voice (171) 833-7472
fax (626) 796-8876 fax (171) 813-1420
http://www.klab.caltech.edu/~geraint [log in to unmask]
-----------------------------------------------------------------
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|