Multinomial model with random effects to study social interactions


I would be very much interested to use a multinomial model with random effects (for a mixed multinomial logistic regression using R).

For the present study, I already analysed independent variables with 2 categories using models of this type:

modele<-glmer(A~ B + C

My independent variable was “A” (with 2 nominal categories).
My two fixed variables were “B” (with 2 nominal categories) and “C” (with 3 nominal categories).
My two random variables were “Ra” and “Rb”.

However, I also have to analyse independent variables with 3 or more nominal categories.
Could you please tell me how I can use “bmrs” for mixed multinomial models? For instance by giving me an example to analyse the independent variable “X” (3 nominal categories) with the same fixed and random variables that I mentioned above.

I thank you very much in advance for your response and your help.
Best regards,

  • Operating System: Windows 7 Professional
  • brms Version: 2.14.4