11-09-2013 04:24 AM
I would like to perform MaxDiff analysis using mixed models with SAS. On surveyanalysis.org I found example of MaxDiff analysis using mlogit package where mixed model was applied. I would like to ask if it is possible to do this analysis using mixed models in SAS (maybe proc mdc, proc glimmix, proc nlmixed or proc mcmc)? I am interested in obtaining random coefficients (solutions) for variable for each respondent.
Link to the example - http://surveyanalysis.org/wiki/Max-Diff_Analysis_Case_Study_Using_R
I also attached input dataset created in section "Setting up the data for a logit model".
Thank you for suggestions and hits how to tackle this problem.