MATLAB as an Environment for Bayesian Computation

The current status of Bayesian software is reviewed. The motivation for the development of Bayesian software is described and the software is categorized by the type of user (student, practitioner, and researcher). The use of the software package MATLAB is illustrated for the different types of Bayesian software. The use of a MATLAB graphical user interface (gui) is demonstrated for the introduction of proportion inference using a discrete prior. A second gui is used to illustrate the use of a MCMC algorithm in logistic modeling with a data augmented prior. The use of MATLAB as a programming environment for the development of MCMC algorithms is discussed, and a MCMC program for fitting a random effects model is outlined. 1 An overview of Bayesian software 1.1 What's available? In a recent issue of The American Statistician, there was a discussion on the desirability of teaching a first statistics course from a Bayesian viewpoint. To see which statistical methods are actually used i..