PREVALENCE ESTIMATION UNDER HETEROGENEITY IN THE EXAMPLE OF BOVINE TRYPANOSOMOSIS IN UGANDA

We examine variance estimators of a binomial parameter established und er cluster sampling using data from a cross-sectional study of bovine trypanosomosis in Mukono County, Uganda. Fifty farms (referred to as c lusters), were sampled with a total sample size of 487 cattle. Trypano somes were found in 17.9% (87/487) of the total sample. The cluster-le vel (CL) prevalences were not homogeneously distributed. According to maximum-likelihood parameters established by mixture-distribution anal ysis, 18% of the cluster had 0% prevalence whereas 48% and 34% of the clusters could be allocated to subpopulations of clusters with mean pr evalences 11.6% and 31.9%, respectively. We show that this form of het erogeneity invalidates the applicability of the Beta distribution as a model for the distribution of CL prevalences. Furthermore, we provide empirical evidence for a variance inflation due to heterogeneity (inf lation factor 2.07) that exceeds the design-based variance inflation d ue to clustering alone (inflation factor 1.82). The variance inflation due to heterogeneity is given in a closed form so that the approach c an be conveniently applied to survey data that involve cluster samplin g under heterogeneity. (C) 1998 Elsevier Science B.V.