I am working on my thesis, where I am using network analysis to describe psychiatric disorders and their bridge symptoms using a network approach. The network is estimated using the IsingFit package.
Common practice in the field is calculating odd-ratios for the weighted edges. I am having trouble computing such odd-ratios, and even more trouble understanding how one would interpret them without their confidence intervals.
Can someone please help ? How can one obtain Odd-ratios for edges in a IsingFit estimated network ? The documentation is very light in regards to this subject.
Thank you to anyone willing to help !
I have tried exponentiating the weighted adjacency matrix obtained in the IsingFit network, which seems to work for OddRatios but I am not sure if that's the way to go. Plus, this fails to give any confidence interval.