This book focuses on Bayesian multivariate inference, which is defined as any inference that involves a multivariate (marginal) posterior distribution and is distinct from frequentist multivariate inference in that that is defined as any analysis with a multivariate response variable.
The text features multivariate models, including the mutivariate normal and t distributions, Dirichlet-multinomial and gamma-poisson models, and generalized mixed linear models. Experimental design techniques are also covered and include blocking and repeated measures; experimental planning principles (Bayesian stopping and decision rules in experiments to prove); and risk assessments such as a-priori Type I and Type II error risks in such experiments.
Decision analysis is presented in this context as a coherent way to compute stopping boundaries and terminal decisions. Planning principles for experiments include information maximization. Each chapter includes problem sets and computer exercises.
.,."simply enjoyable...ideal as a reference." ("Journal of the American Statistical Association," June 2007)
"This book should be very useful in the classroom and for interpretation of the literature, and it is thus highly recommended for students, instructors, and investigators." ("The Annals of Pharmcotherapy," July/August 2005)
.,."a very interesting, well written and...very enjoyable and easy to read book...I highly recommend "Biostatistics: A Bayesian Introduction,."." ("Statistical Methods in Medical Research," Vol. 14, 2005)
.,."suitable for serious students, faculty and researchers, and college libraries supporting biology and biostatistics graduate programs." ("CHOICE," March 2005)