A note on a simple criterion for choosing among sample joints for use in double sampling

Abstract

ABSTRACTA simple criterion is described for selecting which of a number of sample joints estimates most precisely for a given cost (T) the mean lean percentage of a population of carcasses, using double sampling. It is shown that the best joint is the one giving the largest value of:where ρ is the correlation between the lean content of the sample joint and the lean content of a carcass side (y), and k is the cost of the sample joint dissection as a proportion of the cost of side dissection. The optimum proportion of carcasses to side dissect is:and the standard error of the estimated mean is then:where σyis the standard deviation of y and C is the cost of measuring y.