 |
 |
FSL list, We have two groups, patients and controls, and we found group contrasts. We also got a correlation in trauma in all the subjects, now if we want the interactions between groups and traumas, we just need to set EV like this: group EV1 EV2 EV3 EV4 1 1 0 16.5 0 1 1 0 11.5 0 1 1 0 8.0 0 1 1 0 15.5 0 2 0 1 0 7 2 0 1 0 7 2 0 1 0 9 2 0 1 0 9 2 0 1 0 10.5 And for contrasts like this: C1 1 -1 0 0 C2 -1 1 0 0 C3 0 0 1 -1 C4 0 0 -1 1 My question is, do we need to orthogonalise trauma's ev to their groups separately if we only want the interactions? Next question is, from some messages, Eugene said, In many cases, looking for a mean correlation/interaction effect across two different covariates will not be interesting. I agree to this viewpoint. However, I found some papers, in which they reported the interaction between a continuous covariate and groups. Can we try to explore the significance for the interaction? take an example for this one, if we get some activation in C3 , it means controls showed higher correlation with traumas than patients? Perhaps another better idea is, to explore the correlation between traumas and BOLD response within each group separately? Any good suggestions? Thanks a lot, Hongyu