How do I figure out what percentage of health insurance variables were edited in 2019? I'm specifically looking for edited Medicaid coverage and I see two variables in the PUMS Codebook:
I think I should probably use FHNIS4C, but what is FHINS4P? Thanks!
HINS4 is the variable for Medicaid coverage (=1 has Medicaid, =2 Does not have Medicaid).
FHINS4P is the flag for imputed values.
FHINS4C is the flag for values when Medicaid coverage =1 that was assigned as such because of a particular set of characteristics. I think this separate flag is on the file so that researchers can decide whether or not to use medicaid status that was assign due to this set of assumptions.
Essentially, all responses are "edited", but I think you are referring to imputation. Use FHINS4P to get the imputation rate. I don't think there are flags for if values are assigned though a logical edit that is not specific to the "eligibility" edit (such as a logic rule that sets Medicaid coverage to "Not covered").
Edit: I’m adding this link to a good description of the eligibility edit.
Thank you! I am actually interested in both the eligibility edit and imputation. Sounds like I need to look at both fields.