Hi,
I have a question about a design of fMRI experiment. We're interested in
comparing fMRI response to 2 kinds of stimuli (A and B) over the time. In
order to measure the baseline activity over the time and also to equate
the task demand with other previous experiments, we need to present dummy
stimuli explicity as well.
Initially, I was suggested to present 2 dummy stimuli, C and D, and
compare [A - C] and [B - D] (C and D being interchangable) measured at
each block. This would complete orthogonality and would allow us to
compare the main effect of [signal - dummy].
However, our preliminary experiment outside the scanner showed that
using 2 dummies affected task performance, presumably by lowering
saliency of the 2 signals.
So, I'm wondering if I can use only single dummy, C, and compare only [A -
C] and [B - C] over the time. The important point is that our primary
interest is the comparison between A and B. Is there any statistical
problem in comparing [A - C] and [B - C]? Of course, we'll take care of
the presentation order of A, B, and C so that there will be no systematic
difference between the relationship between A and C and that between B and
C.
Thanks
Nao
--
?><?<?><?<><><?<><?<<?<?><>?<<>?<>?<>?<>?<>?<>?<?><>?<>?<>?
Nao (Naotsugu) Tsuchiya
graduate student in CNS program in Caltech
address: m.c.139-74 nao tsuchiya CALTECH, Pasadena, CA, 91125 USA
phone : +1 626 395 2875
fax : 626 796 8876
email in Japanese: [log in to unmask]
homepage: www.klab.caltech.edu/~naotsu
<>??><<>?<>?<>?<>?<>?<>?<>?<><>?<?><>?<?><?><?<?><?><><?<>?<
|