I have 14 groups in the data set. I want to do proportion test for each pair. I used proc freq and that requires a lot of repeated code for each pair. Is there a efficient way to do this? Thanks!
The standard way to compute the same statistic across many groups is to use BY-group processing:
Sort by the grouping variable and then put that variable in the BY list.
Are you looking to make all possible pair comparisons from your 14 groups, ie 14 choose 2?