Mike,
> Do you think it would make sense to use SPM to analyze the original time
> series data for individuals, and then use SnPM for the group analysis
> (random-effects), since the degrees of freedom are limited?
Yes, indeed! This is a typical use of SnPM.
The low degrees-of-freedom case is precisely the case where SnPM can
help out most. By using variance smoothing, you can essentially boost
the degrees-of-freedom and improve sensitivity. An illustration
of this is here
http://www.sph.umich.edu/ni-stat/SnPM/exnew.html
Compare the resulst with and with-out variance smoothing.
-Tom
-- Thomas Nichols -------------------- Department of Biostatistics
http://www.sph.umich.edu/~nichols University of Michigan
[log in to unmask] 1420 Washington Heights
-------------------------------------- Ann Arbor, MI 48109-2029
|