Here's a simple answer to get things started: PROC freq data=have; tables ssn * duplicateid/ noprint out=counts; run; PROC freq data=counts noprint: tables ssn / out=ssnlist; tables duplicateid/ out=idlist; run; proc print data = ssnlist; where count > 1; var ssn count; title 'SSNs having more than 1 ID'; run; proc print data=idlist; where count > 1; var duplicateid count; title 'IDs having more than 1 SSN'; run; There's lots of ways to slice and dice the numbers. This is just a reasonable start.
... View more
