Four examples are given of how discussion between scientist and statis
tician enhanced both the analysis and interpretation of experimental d
ata. They cover both the design and analysis of field experiments, and
some of the problems associated with interpreting data from dose-resp
onse assays. There is little doubt that the statistician, with his arr
ay of analytical and diagnostic tools, can often improve a scientist's
interpretation of experimental data, although a statistician cannot c
ompensate for a poorly designed or inadequately replicated experiment.