Some practical considerations in sampling livestock populations to estimate disease prevalence and other parameters

Inferential statistics consists of methods for making inferences about parameters of a populations, such as the disease prevalence in a livestock population, based on estimates of the parameters obtained from a sample of the population. Since the sample size is usually much smaller than the size of the population it is imperative that the sample be representative of the population sampled. This paper discusses two general characteristics of estimates that are used to evaluate the accuracy of an estimate of a population parameter and presents the general class of sampling designs that can be evaluated in terms of these characteristics. Examples of sample designs that belong to this general class are described with emphasis on those designs that would be practical in sampling populations for estimating the prevalence of disease.