I am attempting to determine respondent's mother's age at the respondent's birth using: AGE_MOM - AGE
However, this results in implausible values (range = -4 to 78), for example the mother was 78 when giving birth or the mother is even younger than the child. Is this a data problem, or is there a more appropriate way to construct the variable?
Note: AGE_MOM was from the IPUMS download.
Thanks,
Deepa