Hi all,

I am in the midst of setting up my data to run a panel-data mixed logit choice model (cmxtmixlogit) in Stata16.

Is it possible to specify either of the following when estimating?:
  • An attribute that is only relevant for a specific alternative
  • A no-choice alternative (i.e. a base alternative if a respondent in the choice experiment opted for non of the presented options)
After reading the relevant documentation, it seems that Stata can only handle choice tasks where all attributes appear against all alternatives, hopefully I am just missing something.

Thanks in advance for your help.
Paul