Hi Victor,
> However, I would like to perform the same F tests for the main effects
> and interaction with *SnPM*, and here I've
> got stuck, because SnPM always splits an effect into a negative and a
> positive part.
If you just want the two-sided property of F-contrasts, then there is
an approximate way to get this. From the SnPM manual,
http://www.sph.umich.edu/ni-stat/SnPM/man.html
A two-sided test may be constructed by doing two separate
analyses, one for each tail, at half the chosen significance
level, doubling the resulting p-values.
(Strictly speaking, this is not equivalent to a rigorous two-sided
non-parametric test using the permutation distribution of the
absolute maximum statistic, but it'll do!)
If you want matrix F-contrasts rather than just two-sided t-contrasts,
then I think SnPM5 supports these, based on a quick google and a
glance at:
http://www.sph.umich.edu/ni-stat/SnPM/snpm5b_updates/snpm_cp.m
Best,
Ged.
P.S. FSL's randomise (similar to SnPM) promises F-contrasts in the
next release, due out in a month or two, I believe...
|