When you run tests on large datasets, there can be times when the results become skewed due to unseen errors to the procedures. This is one of those times.
I had one of the project scientist look over the data and procedure. They found a problem with how the data was compiled into the individual networks. Some of the links for the character networks combined improperly and resulted in larger link weights. This, in turn, altered the results.
I am working on how to proceed with the corrections. But that means the previous analyses will need to be rerun and new results will be posted.
I was told that I am not alone in this.
I have begun correcting the data and will start a new set next week. This is a small setback but I am doing all I can in order to correct the problem.
I thank you for your patience.