Graph theory after 1st level regression

Hi Alfonso,

I posted my question in another thread but I guess it's too long a thread to keep track of the messages, so I'm posting it again as a new topic.

I have run a bivariate regression analysis on resting fMRI data across 11 subjects. Since the resulting network represents the 'effective connectivity', I would like to apply graph theory analysis on this resulting 'effective brain network' to study it's organization. But in the network explorer, the network edges can be selected based on cost, correlation coefficient and z-score. I would like to know what these would interpret to when run on a graph resulting from bivariate regression analysis. For eg. would the z-score mean the z-transformed value of the slope of the roi-roi bivariate regression line? Likewise what would correlation coefficient in this case be interpreted as?

Also, regarding the 'results explorer' in the second level analysis, as someone had mentioned in another thread, is there a possibility to visualize it based on the significant beta values instead of significant T-values? I am using conn 15.h.

Thanks
Shukti