I have imputed my data and there come in some negative values in Income variable and in some categorical variables. What should be done of that? please guide I ll be very thankful. Also the constraint tab is not allowing me to put min and max values. Is there any way ? all the variables are categorical.
Hello Zoiya,
Obviously, whatever method you are using for imputation is presuming that the target values are continuous and unrestricted in range. That obviously isn't in keeping with categorical variables. Choice of an imputation method can and should be conditioned on the nature of the variable under consideration.
You didn't indicate whether your variables, being categorical, are properly classed as nominal or ordinal. There would be a big difference in imputation methods between the two scale types.
If you are convinced that your chosen method is defensible and appropriate for the task at hand, and the resultant values are intended to be categorical, you'll have to figure out a conversion scheme that is also defensible for your specific situation (for example: all "negative" imputed values are assigned to the lowest category for a variable). Given the way your query is worded, though, I would urge you to consult with someone at your institution who has more experience with handling missing data for guidance on how to proceed.
Good luck with your work.
- Badges
- Science topic
- Similar topics
- Chemistry
- Pharmacology
- Pharmaceuticals
- Drugs