http://www.amstat.org/publications/jse/v14n3/langford.html
Gosh - I did not realize there were so many different ways to calculate percentiles! I just noticed my results from Excel (formula =percentile (array, percentile) and JMP (quantile plots and data) differed, and that led me to the above website, which identifies about 12 or more different methods! I am at a bit of a loss, for I often use results from both programs (obviously that has to change). Or I calculate percentiles in Excel when summarizing data and run stats in JMP later - which means two different methods were used.
The above publication recommends one alternative that is given in SAS, which I believe is a bit different than the JMP method.
I suppose the best solution would be if there was a way to make the same calculation in JMP and excel, since those are the programs I mostly use. Is there?
Thanks in advance.
