Dear all,
I would like your assistance in performing and interpreting three-way interactions in SPM. In a current study we have three continuous variables and we would like to study the effects of these variables on resting state networks while controlling for the others. We would like to test for the main effects of each variable as well as the interaction effects. We understand that in this model we would have 7 effects in total:
- 3 main effects (one for each variable)
- 3 two-way interactions (one for each pair of variables)
- 1 three-way interaction (interaction with every variable)
We already performed an analysis in which we included 7 regressors: one for each continuous variable, one for each two-way interaction that was the multiplication of the corresponding continuous variables and one for the three-way interaction that was the multiplication of the three variables.
However we are not sure this is the correct way to perform this kind of analysis (specially in the three-way interaction) and how to then dissect the interactions. We would kindly appreciate any assistance.
Best regards,
Paulo Marques.
|