Hello again Tom et al,
A couple of further issues came up regarding smoothing for the single group randomise analysis.
- My inputs for randomise are copes from a FEAT analysis. These have already been smoothed at 5mm FWHM. When I then enter them into randomise, am I in effect "double smoothing"?
- In the "help" doc for randomise, I see the following:
-v <std> use variance smoothing (std is in mm)
so it seems as if I do choose to use smoothing (pending the point above), I can use mm after all...?