help > Cluster-level connection is significant, single connection isn't
Showing 1-1 of 1 posts
Display:
Results per page:
Jun 5, 2024  10:06 AM | Riccardo Bonafini
Cluster-level connection is significant, single connection isn't

Hi CONN users and creators, 


I'm working on my thesis project using CONN and came across a strange result. I ran an RRC analysis using 8 anatomical ROIs and computed the results by comparing two conditions. Here, I got a cluster of connections (with one connection only) that reached significance. However, when I explore the single connection (the only one present in the cluster) it shows a p-FDR > 0.05 (0.085). 


My questions are: 



  1. How should I deal with this issue? Would it be better to focus on the cluster-level statistics and forget about the single connection?
  2. How is it that a cluster of connections has only one connection?

 


Thanks in advance!