Hello, I have a data set that contains survey results about managers. For each manager we evaluate the percentage of responses that were a "strongly agree" . This is a calculated measure/variable in my report. I need to break the managers into three groups depending on their percent strongly agree ([SA]) - so, High, Med and Low groups. I've tried using the percentile function to figure out the scores to base the breaks on, but in a table that contains manager name and percent strongly agree, when I add Percentile([SA];.33) it always returns whatever the value of [SA] is. I've tried to change the context and still have not yeilded any meaningful results. I have also tried to use the Rank() function and then to break them into three even groups, but have been unsucsessful because the rank function does not rank in order - for instance if you have two scores that tie, the next number in sequence is not what I would have expected: 1,2,3,3,5 - notice the 5 not the 4. So if I were to take the top number of a rank in series, and divide that by 3 I'm not going to get an even distribution because the rank numbers are not in sequence. Can someone please offer any advice?