THE INVITATION to appear at this symposium provided an opportunity to review the changes which have occurred in the relationship of the statistician and the clinician. Only a few short years ago in addressing a group similar to this one, I would have stressed the technical aspects of statistics. I would have amused myself and the audience by reciting horrible examples of false conclusions from then current literature. The situation has improved considerably over the years. Workers in clinical research acquired some knowledge in technical statistics. Often they beat the statistician to the punch. They even have learned to calculate P.
Many of us are pleased with this change and happy to relinquish the position of watchdog of statistical validity. In fact, the time is not far off when the tables will be turned, and the clinician and biologist will monitor statistical investigations for biologic validity.
In many mathematical and