Hello there, Thank you for the quick response. I have been trying to get the 1:1 matching based on (1)performance (ROA) ( within 30% of absolute difference), (2) the same industry and (3) year. For example, I created the Sample (5000 obs.) and the Totalcontrol (30,000 obs.) based on my dummy variable "Suspect"; Each observation has an id, year, industry (sic), ROA, and other variables. I followed the proc sql; create table potentialmatch1 as select * from sample, totalcontrol where sample. sic = totalcontrol.sic; quit; It seems that I did not get it right. The output "potentialmatch1" has over 8 million observations. Did I get a wrong start? I mean do I have to rename the variables in the Sample so that these two datasets have different variable names? Thank you very much. I really appreciate it.
... View more