The Statistical Treatment of Hatchability Data

IN experimental work with poultry, investigators are often confronted with the problem of determining how well the fertile eggs produced by a given lot of birds will hatch. Some statistic, which will characterize the hatchability of the fertile eggs produced by the entire lot of birds, and an estimate of the standard error of that statistic are required in such instances. This paper presents a discussion of two such statistics, based on two distinct points of view regarding the kind of information to be extracted from a given set of data, together with methods for computing valid estimates of their standard errors. The first of these statistics, which will be denoted by p̄, may be described briefly as a weighted average of the hatchabilities computed for individual birds in the given lot. The weight which is assigned to the hatchability for each bird is found by making use of the . . .