Dear everyone,
One note on AlphaSim / 3dclustsim in combination with SPM, I have the feeling that it is often used incorrectly, based on the FWHM of the smoothing kernel during preprocessing instead of that of the residuals, with the former usually being smaller than the latter, or mixing up smoothness in voxel units vs. mm, or using an incorrect mask file reflecting the spatial resolution of the raw data and not that of the preprocessed data. At least the reported extent thresholds are often surprisingly small, and if you search a little, you'll find various threads in which people started with incorrect settings.
Which doesn't speak against the method as such of course. But as soon as you switch between platforms you might easily run into problems in case you're not familar with the default settings/assumptions of the other toolbox.
Best
Helmut
|