ApolloChoiceModelling forum

Dear Stephane and David,

Thanks for your incredible work on Apollo!

Sometimes there are alternatives in a choice data set with just a few choice observations, so called rare events.
Does it make sense to use penalization to obtain firth bias corrected estimates, like available within JMP software by SAS?
And how can I obtain these firth bias corrected estimates in Apollo?

Thanks in advance!

Kind regards,

Toon

Hi Toon

we've looked into this, and the kind of correction would be impractical to code in a general way (so it can be used with all models) and would also only appeal to a limited number of users of Apollo.

Best wishes

Stephane