Dear SPM list,
I want to use a first-order parametric modulator (zygomaticus EMG), in the 1st level model, which adds an extra regressor in the SPM.mat next to each modulated condition. Importantly, I am actually interested in the modulation of the BOLD response by the PM, and don’t just want to control for the PM in order to get rid of its effect. Moreover, I expect the PM to specifically modulate the BOLD of one condition more than in the other condition.
My questions are:
Are the values of the PM automatically centered (demeaned) and scaled (z-scored) by SPM?
If they are, does this happen separately for each session and condition? Based on what I found in the SPM list so far
(e.g. see https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind1305&L=spm&D=0&1=spm&9=A&J=on&K=2&X=01422C62AD656C4D8F&Y=skorb%40wisc.edu&d=No+Match%3BMatch%3BMatches&z=4&P=249890),
it seems like they are at least centered for each session. But the question remains if they are also scaled separately per condition, since I am afraid that this might reduce the potentially different impact of the PM on each condition. do you know if SPM does automatically scale the PM for each session and condition, and if yes is there a way to prevent this from happening?
I wonder about what is the best contrast to test my hypotheses, and whether some of the possible contrasts with the PM make sense.
Specifically, I am interested in the BOLD activity in condition A > B relating to the PM. My plan is to look at A*PM > B*PM.
But what about the contrast (A*PM > A) > (B*PM > B) ? Is one even allowed to contrast the regressor with PM and the regressor without PM ?
Finally, based on the result, it looks to me like the contrast A > B is identical to (A + A*PM) > (B + B*PM), but how is that possible, since A*PM > A and B*PM > Y give clear effects?
In relation to this, a 2011 comment by Jonathan Peelle on the list
(https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind1107&L=spm&D=0&1=spm&9=A&J=on&K=3&X=3CE13241B91BCB0A20&Y=skorb%40wisc.edu&d=No+Match%3BMatch%3BMatches&z=4&P=188475)
states:
“The parameter estimate for A reflects the effect of A that can't be
explained by the parametric modulator, and the parameter estimate for
AxPM explains the effect of the parametric modulator that can't be
explained by A (just as it would be with any two regressors in the
GLM). However, since the parametric modulator is mean-centered on the
main effect, these effects shouldn't be confounded anyway (i.e. the
mean of a condition doesn't depend on how the response varies as a
function of some other factor).
Does this mean that if I am interested in the BOLD activity related to the PM in condition A > condition B, I should first look at A > B, and then use the outcome as an inclusive mask for the contrast A*PM > B*PM ?
Your help is greatly appreciated.
-----------------------------------------
Sebastian Korb, Ph.D.
Department of Psychology
University of Wisconsin, Madison
|