hi everyone (especially Gang)
I have a question about using 1dSVAR. We have a design where participants watch 3 types of video clip, A, B and C. The idea is to identify a network using the A videos (using PPI or 3dGC.R) and then seeing whether certain connection strengths between regions in this network differ while watching clips of type B or type C. Would it be statistically justifiable / 'kosher' to run 1dSVAR on the B trials and the C trials separately and then test the difference in connection strengths using a paired t-test? or am I way off base here?
thanks in advance
James