The attached subprotocol implements a scatter plot by category with outlier detection, jitter of points, an associated table of summary statistics for each group, and calculation of t-test for each group against the identified control group.

It is designed specificly to work with data from in vivo or clinical studies, but can be used with any data (as shown by the example protocol which uses the automobile test data set).

This subprotocol is free to use, but comes with no warrantee. If you make improvements please post them back to this forum.

Leo S. Bleicher
Cepheus Information Systems
http://www.cepheusinformationsystems.com/