Hi,    I am looking at a data set of teachers who are combined teaching 129 different courses across  the internet.  In an effort to guarantee the quality of the courses, the 12 teachers should be graded.    Each month, 10 % of the courses offered should be randomly be chosen and graded, whereby the teachers teaching these courses should be evenly distributed over the sample taken. Additionally, the courses that were sampled last month, should not be sampled again.    Proc Surveyselect data = teachers_&end_current-mth out teacherevaluation_&end_current-mth method = SRS seed= 12345678 Samplerate= 0.10; run;   This code, gives me 10% randomly chosen samples of teachercourses to be sampled -  without repetition. So far so good, but some teachers are more often sampled than others.   What condition can I add to this code to make sure, the samples randomly chosen are evenly distributed across all teachers?  Your help is much appreciated.  Many thanks, Anna     ... View more