Think of a specific problem of comparing two treatments preferably from your own field. How would you organize an experiment to find out which one was better? What are some of the problems that might arise? How would you analyze the results? How could past data be used to assist in choosing the design and the number of runs to be made?