Dear CONN users,

I recently found out that the order with which you organize the brain regions of your Roi to Roi analysis may hide the results

So if you are performing an ROI-to-ROI analysis and obtain an empty connectome as a result—that is, no connections survive the FDR- or FWE-corrected significance threshold of p < 0.05—that is not necessarily a negative result.

After trying three different ROI orders and obtaining only empty connectomes, I finally found a meaningful and statistically robust result by loading a fourth Roi to Roi region order. 

I believe by the always fast improving CONN updates, Alfonso et al will soon provide a way to find this sweet spot.

All best!

Renzo