help > Multivariate Stability Analysis
Jun 7, 2018  08:06 PM | Jason Smucny - UC DAVIS
Multivariate Stability Analysis
Hi Conn Experts,

Thanks for your wonderful toolbox. I was wondering, is there (or will there be developed soon?) a straightforward method of comparing connectivity matrices longitudinally to test for stability/test-retest reliability? In other words, given a sample of N controls and Y patients with baseline and follow-up connectivity data, do you know of any good ways to do a multivariate analysis to compare the stability (maybe interclass correlation coefficients?) of the overall matrix between controls and patients (and then follow that up by analyzing stability of networks within that matrix)? I've heard of methods like Structural Equation Modeling, although I have not done such an analysis and don't know how straightforward that would be to implement.

Thank you,