SUMMARY A series of trials of new varieties of a crop may be undertaken for one or more of several reasons. Among these reasons are: comparing the new varieties with a standard, testing differences between new varieties, selecting the best varieties, and estimating true yields of selected varieties. The replication (or even the inclusion) of a standard variety should depend upon the emphasis to be placed on these reasons. A model that can apply to all these situations is discussed. For most purposes, it appears that the standard variety should be replicated 2 or 3 times as heavily as any one new variety; this factor should rise towards the square root of thle number of new varieties if emphasis lies chiefly on comparing each new variety with the standard and should fall to 1 (or, in extreme instances, 0) if emphasis lies chiefly on selection.
No takes yet. Share an insight, caveat, or question.
David John Finney (1964) studied this question.