ApolloChoiceModelling forum

When incorporating income into a utility function, I encounter a challenge due to missing income data from some respondents, as they are unwilling to provide this information. How should I address this issue of missing data? Should I replace it with 'na' or other values, or do I need to remove that observations?

Please advise

Hi

you can either impute them, or estimate a separate effect for those people

Stephane