Yes. I would recommend these "true" baselines be
explicitly modeled. You might realized that within the event block you
might not be able to figure out what the baseline is. Thus, if you
explicitly model these baseline conditions, you might have better
chance to reach the contrasts you would expect.
On 4/13/07, Saemann Philipp <[log in to unmask]
we have a rather basic question on how breaks within an event-related fMRI
paradigm (four types of "events") should best be handled.
The final goal of the analysis is to contrast the four different types of
events against each other.
After block-like presentation of the four different, intermingled events,
breaks of 20 seconds are programmed to allow
the subject to rest. Should this "lower" baseline be explicitly modelled in
a block like 5th regressor, or does this just
introduce some collinearity that is not wanted?
Thank you very much for any advice,
Max Planck Institute of Psychiatry
NMR Research Group
Mail: [log in to unmask]
Jeng-Ren Duann, PhD
SCCN, Institute for Neural Computation
University of California San Diego
4150 Regents Park Row, Suite 320
E-mail: duann at sccn.ucsd.edu or
duann at salk.edu