Groups of Rhode Island Red chicks were fed a mash made up from East African feedingstuffs or that diet supplemented with a high level rate of aureomycin—97 g. aureomycin hydrochloride per ton of feedingstuff. The chicks were kept on wire for 8 weeks. The initial improved growth rate of the chicks fed the aureomycin supplemented feedingstuff was not maintained, nor was there any statistical indication that the addition of aureomycin improved the feed conversion rate.

Although the mortality rate appeared to be very much lower in the chicks fed the aureomycin supplement, the evidence was too limited to allow this to be attributed solely to the effects of the aureomycin.

The experiment illustrates the need for careful design in order that the effect of the factors to be tested may be separated from the effects of other factors contained in the experiment and not confounded with them. Moreover, advanced statistical analysis should be used as an invaluable aid to, but not a substitute for, visual or graphical examination of results, for such examination often yields valuable indications of possible lines of analysis and may avoid the drawing of unjustified conclusions.