Sampling Intensities and Replication Requirements for Detection of Treatment Effects on Testicular Function in Bulls and Stallions: A Statistical Assessment

Publisher Website
Google Scholar
Abstract
Data from testes of 16, 2- to 3-yr-old stallions and 34 yearling beef bulls were utilized in a components of variance approach to calculate the number of observations required per testis and(or) the number of animals required per treatment group to provide experiments of known sensitivity and precision, where treatment was to be assessed by one of several endpoints. The latter included paired testes weight, seminiferous tubular diameter, the number of germ cells per seminiferous tubular cross-section, or the number of elongated spermatids per gram of testicular parenchyma or per testis. For all variables for which several observations were available for each testis, precise assessment of any given male required far fewer observations than have been used routinely. However, replication requirements varied substantially in relation to the sensitivity (i.e., size of difference to be detected) and precision (i.e., Type I and II error probabilities) desired. Replication requirements were greater for stallions than for bulls, particularly at higher levels of sensitivity, for which requirements for both species were very large. The data presented should permit future experiments involving assessment of these endpoints to be designed with known sensitivity and precision and with optimal efficiency and cost-effectiveness. Copyright © 1989. American Society of Animal Science . Copyright 1989 by American Society of Animal Science