Thomas E Nichols wrote:
> The short answer is 'yes', similar problems would be expected.
> Generically speaking, non-parametric permutation methods are nothing
> more than a way for getting more trustworthy P-values. Problems of
> roving peaks in your t images won't be fixed if you just change your
> P-value threshold.
Thanks Tom.
> More subtly, though, SnPM allows you do use variance smoothing, which,
> under low DF (say, < 20) will regularize your variance and make your
> t-image more like your difference (contrast) image. This might make
> for more satisfying results (it certainly seems to improve power).
Will this have virtually no effect with say 50 images?
On a faintly related note, how do the capabilities of FSL's
"randomise" compare to those of SnPM?
> My own $0.02 on your initial problem, though, is a question: Why were
> you looking outside the brain in the first place?
Essentially because I'm interested at this point in the methods
themselves, e.g. in the differences between spm2 and spm5, and in the
validity/explicability of the basic results -- with as little in the
way of masking/thresholding/tweaking as I can have. Also, in this
context I am more interested in the t-values than I am the corrected
p-values.
I am also helping with some more clinical projects, where the goal is
very different -- there we have traditionally used tight explicit
masking, and will probably continue to do so.
I really appreciate your comments, thank you,
Ged.
|