Hello, 

I have completed a seed-to-voxel gPPI analysis (regression: bivariate) for my task-based functional data(within group). I then extracted the significant clusters for my task contrasts. I am trying to run further analyses with these connectivity measures but it seems like they are not very sensitive to the other predictors that we have in our dataset. 

I was wondering if anyone would have suggestions on how to make these values more robust. Would it be appropriate to use methods like factor analysis or PCA on brain connectivity data?

If anyone has any suggestions I would appreciate it. 

Thanks in advance! 

Didi