@SBuc wrote: is there any advices or SAS tutorial on how to perform multiple regression analysis with Proc MCMC and especially for dealing with variables that are crossing the no-effect cut-off and final variable selection? Google finds a number of links that might be useful. As I have never done this, I can't be more definitive. ... View more

You can try an ordinal regression.  The documentation has a fully worked example:   https://documentation.sas.com/?docsetId=statug&docsetTarget=statug_logistic_examples03.htm&docsetVersion=15.1&locale=en   Under the title there's a link to the full code that you can run on your machine and then follow through. Once you understand the example, you can replicate it on your own data.    @SBuc wrote: Dear Brain trust, I want to determine the effect of a treatment on a score of pain in animals. Many published literature has used GLM models using score as a continuous dependent variable accounting then for treatment, block and animal repetition using linear mixed model. however, this approach is per se assuming that the score is a continuous variable (which is not) and also that a score of 8 is 2 times worst than a score of 4 which is not demonstrated (so the interval between 2 unit scale can not be assumed to be the same throughout the scale (0-15 max points). initially we used non parametric tests such as rank comparison but I want to know if a more elegant and powerful analysis can be performed. thanks   ... View more

I had had the same thought as @PGStats initially, but I wasn't sure I was interpreting correctly.   If age is a fixed effects factor that "causes" enzyme concentration (a random variable), I would still use a random coefficients model. You can then use that model for "inverse prediction", also known as "inverse regression" or "calibration".    I cannot say that I've ever attempted that with a random coefficients model, so at the moment the best I can offer is a few promising links:   Inverse Prediction Using SAS® Software: A Clinical Application   Approximate Graphical Methods for Inverse Regression   On Inverse Prediction in Mixed Linear Models   investr: Inverse Estimation/Calibration Functions  (R not SAS!)   I hope this helps.     ... View more

An ROC analysis can be done by using the predicted probabilities from the GLIMMIX model in PROC LOGISTIC as discussed and illustrated in this note.  ... View more

Thanks for this reply. In fact we have 55 farms where we have at least 12 or more calves (up to 20). we have an extra 20 farms where we have from 5 to 10 calves. since our inference would be at the farm level it is why we want if possible to stay at the farm level for all our analysis with the idea that what is good at the farm level is not necessarly exactly the same at the calf level. ... View more