I've read a lot about psychophysiological interactions in
factorial studies designs, my question is how one extends some
of these issues to simple categorical designs.
For example, if one has PET data of subjects in two states (A,B)
would the following be a sufficient protocol for studying
psychophysiological interactions:
1. Determine the main effects (A-B)
2. Select a region of interest based on these main effects.
3. Extract metabolic activity within that region using the VOI
routine (adjusting for confounds of no interest)
4. Use the extracted activity as a covariate in a "single
subjects conditions and covariates" design(SPM99)
5. Select an interaction between the condition (A,B) and the
covariate such that there will be 5 columns of the design matrix
(2 state columns, 2 covariate columns, and 1 global column)
6. Assess the interaction using contrasts: [0 0 1 -1] and [0 0
-1 1].
Furthermore, I assume if these states are manipulated with a
drug these interactions would more specifically be refered to as
pharmacophysiological interactions...right?
Thanks a bunch!!
Nate White
=====
Nathan S. White
Staff Research Associate
Department of Anesthesiology
University of California, Irvine
[log in to unmask]
__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com
|