Publisher’s description: Assuming a basic knowledge of the frequentist approach to finite population sampling, this monograph begins with a discussion of the underlying foundation of the subject. It describes the Bayesian and predictive approach to inferential problems in this area with an emphasis on the likelihood principle. Bayesian Methods in Finite Population Sampling then gives a non-informative Bayesian justification for standard frequentist methods and studies extensions of the Polya posterior. The problem of the simultaneous estimation of several stratum or area means is considered and the book develops both the empirical and hierarchical Bayes approaches for these problems. Models which include covariates are given and their relationship to general mixed linear models is studied. Appealing primarily to graduate students and researchers in finite population sampling, the book has also been written for Government department statisticians who use sampling and will be of great interest to lecturers and researchers in statistics and biostatistics in general.