Sampling a hierarchical ordered logistic regression

Questions
3

Hi AlexAndorra,

  • Have you tried with different priors to see if this affects your prior predictive checks?

I tried adjusting the priors a bit but didn’t see any notable differences.

  • Sampling and ADVI issue: have you standardized your predictors? Predictors that are on a very large scale often mess up MCMC sampling.

The predictors have values between 0 and 10 so I didn’t think those scale issues would apply here.

  • You could find this notebook useful – it’s an adaptation to PyMC3 of McElreath’s Rethinking chapter 11 (first edition).

Thanks. I think my code is similar to that code, but the hierarchical aspect in my model does complicate things.

I tried using an OLS instead of the ordered logit and that seemed to work better – at least it converges. I can stick with OLS for now, but if anybody has suggestions on how to make the ordered logit work I remain interested.