I'm bootstrapping samples (with replacement) for a matched case control data using proc survey select. The sampling unit is the id for each matched pair and am outputting all the replicates using the outhits option. So the output data set is the same size as the input data set.   My goal is to assign different id values for each duplicated matched pair to prevent such case-control pairs from being dropped off during analysis. Here is an sample input and output data set, and the modification am trying to achieve.    I'm thinking by group processing but when the number of hits is more than 2, then I can't seem to wrap my mind around it.   Any ideas are appreciated.      ... View more