Dear SPM list,
I wish to analyse the effects of one parametric modulator (Pmod) on the HRF informed basis set (i.e. three basis functions) in two independant subject groups. My problem regards 1) the contrasts best suited to answer my research questions and 2) the interpretation of such contrasts. Any guiding advice will be much appreciated!
At the 1st level, individual design matrices included condition regressors (A) and parametric modulator regressors (AxPmod), both convolved with the three basis functions (i.e. A1, A2, A3, AxPmod1, AxPmod2, AxPmod3). Basic t-contrasts on the AxPmod regressors resulted in three con images for each subject.
At the 2nd level, I used a 3x2 Full factorial design with Basis set as factor #1 and Group as factor #2. Now, here is where I need guidance :
The following F-contrast :
1 -1 0 0 0 0
0 0 1 -1 0 0
0 0 0 0 1 -1
would answer the question : "Is there a group difference in the modulation of the shape of the HRF by Pmod?", but would not inform the nature of the difference; e.g. in a given significant volume, the correlation between Pmod and betas is
a) positive in Group 1 and negative in Group 2?
b) more positive in Group 1 than in Group 2?
c) null in Group 1 and positive in Group 2? etc...
Am I correct? How can I assess the directivity of the slopes to interpret the results?
Thank you very much for addressing this issue!!!
==
Nicolas Martin, Ph.D. candidate
Université de Montréal
Center for advanced research in sleep medicine
Montreal (QC), Canada
|