Dear Statalist,
I would like to know ways to use Stata to simulate a continuous distribution of an income variable that is censored (i.e., respondents of the survey do not declare directly their income, they can only choose options such as "Less than 15000", "From 15001 to 30000", etc.). The reason I'm doing this is that I would like to obtain the total wealth of an individual (including their asset etc.) and that only the income variable is censored, which is not convenient as I have to sum all components of the individual total wealth. Please note that the income variable is only censored in one wave of the survey, while in others it continuous.
I know about the command -intreg- that can be used to predict variables based on a range of predictors, but if possible I would like to expand my knowledge about other things that can be done in Stata to achieve this goal. Paper suggestions that deal with this issue are also welcomed!
Best,
Adam
I would like to know ways to use Stata to simulate a continuous distribution of an income variable that is censored (i.e., respondents of the survey do not declare directly their income, they can only choose options such as "Less than 15000", "From 15001 to 30000", etc.). The reason I'm doing this is that I would like to obtain the total wealth of an individual (including their asset etc.) and that only the income variable is censored, which is not convenient as I have to sum all components of the individual total wealth. Please note that the income variable is only censored in one wave of the survey, while in others it continuous.
I know about the command -intreg- that can be used to predict variables based on a range of predictors, but if possible I would like to expand my knowledge about other things that can be done in Stata to achieve this goal. Paper suggestions that deal with this issue are also welcomed!
Best,
Adam
