Hi all!
For my CBC, I have to report the RLH (Root Likelihood) alongside the Percent Certainty (Pseudo R2) as a quality criterion for internal validity.
I find both values in the output of the logit. I was wondering about the RLH: what is the difference between the RLH values from the HB, which I get separately for each respondent and could theoretically average(?), and the RLH from the Logit. According to my data, they are also quite different: averaged RLH from the HB 0.701 and RLH from the logit 0.4131.
In many papers they always just briefly give the number to the RLH (sometimes it is called average RLH), but never exactly how they did it or on which values it happens. When the authors talk about RLH, presumably they mean the RLH from the logit? Or are they averaging the RLH from the HB (if one is even allowed to do that? - I haven't found anything on that).
I would be grateful for any help