MEASURING YOUR POSTERIOR BELIEFS THAT YOU CAN COUNT PROPERLY

In toxicity studies with dogs, animals are dosed on an individual milligrams per kilogram basis, adjusted weekly For practical reasons the experimental compound is manufactured as tablets of different strengths, so that individual animal doses are made up of varying numbers of tablets of these different strengths. In this paper a Bayesian approach to determining the likelihood of counting errors encountered in preparing the dose for individual animals is investigated. A Bayesian approach is particularly appropriate in this context as we shall have strong beliefs in the ability of individuals to count small numbers of tablets correctly, but weaker beliefs in their ability to count larger numbers.