I have difficulties identifying the point when convergence in HB estimation is established:
Although some kind of leveling starts after about 3'000 iterations, parameters still vary a lot.
Is there a possibility to evaluate convergence by analysing the saved betas of the parameter from one of the files in the advanced output options, i.e. Successive estimates of alpha?
If so, would be looking at the development of a parameter's variance of alpha/beta be a way to look at it?
Thank you in advane for your help