Hi

I am using Crystal 9 using Access 2003 database as a data source.

SynapseVampire responded once to a query about the differences between Excel and Crystal's percentile calculation with the following:
1 - Crystal does NOT interpolate bewteen low and high values where the rank is not a whole number. For example,if the target rank position for percentile P is 6.25, the interpolation of the Pth percentile **should** be [datapoint 6] + ([datapoint 7]-[datapoint 6] * 0.25). Crystal, however, takes the simple average (this seems to me like nothing more than laziness on the part of the CR development team....)
2 - when determining the rank position for percentile P, Excel uses the algorithm [1 + P * (count_of_datapoints - 1)]. Crystal, however, uses [P * (count_of_datapoints + 1)]. Not sure about the whys and wherefores of this one (the NIST page I reference for stats info is down at the moment so I can't dig into it any more).

Using the above as a reference how would I get the 90th percentile for patient lengths of stay in a crosstab?

Thanks.