Dear everyone,
I'm planning to test the updated version of the SnPM toolbox, as it seems to be comptabile with SPM12 now. Basically I want to run some within-subject ANOVA on a data set with relatively few subjects (12 subjects).
Looking at the provided fMRI example http://www2.warwick.ac.uk/fac/sci/statistics/staff/academic-research/nichols/software/snpm/man/exnew/ I noticed that a initial threshold of .0001 is suggested. Now the thresholds are arbitrary of course, but is there any specific reason why it is .0001? And not for example .001, which is quite common as an initial uncorrected voxel threshold for parametric SPM analyses, followed by a corrected cluster threshold.
Best,
Helmut
|