[VoxBo] Comparing two contrasts

[David January]

Hello,

I want to compare two contrasts to see if there are any brain regions
more responsive to one than the other.  Said another way, I ran a
study with 4 conditions.  I've tested for the difference between 1 and
2 and the difference between 3 and 4, and now I want to see if (1-2)
is greater than (3-4).  Is there a way to do this in VoxBo?  I stole
this comparison from another paper and they say they did the following
(though I'm not committed to doing it this way if another way is just
as good):

For each participant, the Kolmogorov-Smirnov statistical test was used
to generate statistical maps for the comparison of [1-2] vs [3-4].
Stastical maps were then converted to Z-score maps and transformed
into a common stereotaxic space [I normalized before analysis, so I
don't need to do this step].  A map was generated in which each
voxel's score indicated the percent of subjects for which the Z-score
exceeded 2.32 (P<.01).  Voxels showing activity in a majority of
participants (9/16 or better) were then selected for generation of
activation maps (according to the binomial distribution, for an n=16,
the voxel-wise probability for falsely meeting this criteria is
~1x10e-14).

This whole enterprise is embedded in a mixed event-related and blocked
design, so there's a chance I'm misinterpreting what happened.  The
blocked aspect of the design is that there was a block on only 1 and
2, a block of only 3 and 4, a block of only 1 and 2, and a block of
only 3 and 4.  The event-related part of the design is that within
each block, the order of conditions was event-related.  The actual
text of the paper (Milham et al., 2001, section 2.8, p. 469 in
Cognitive Brain Research, vol 12) refers to the condition labels for 1
and 3 where I put [1-2] and [3-4] in the above paragraph, but I was
interpreting that as a shorthand label for the blocks because it seems
you wouldn't need to do all this fancy math to compare 1 and 3
directly - would you?

-- 
David January
Department of Psychology
University of Pennsylvania
http://www.sas.upenn.edu/~djanuary/