Bayesian statistical methods are applied in situations of decision making with limited information in diverse fields including economics, law, medicine, psychology, and meteorology. The objective of the Bayesian Analysis, Computation and Communication (BACC) project is to provide the wider social science community with rapid, convenient access to state-of-the-art simulation methods for Bayesian analysis, computation and communication. To meet its objectives, the project will develop software incorporating recent innovations in mathematical statistics and computer science. A unique and important feature of the BACC software is that it implements its tools as extensions to the popular mathematical applications programs Matlab, Gauss, S-plus, and R. The architecture that makes this feature possible is unique within the domain of Bayesian software, combining core code, model-specific code, application code, and a specialized mathematical routine library.
This project will expand the suite of models provided to users, provide a set of high-level commands for users developing new models, and incorporate structured procedures to cross-check the validity of algorithms. It also will emphasize outreach and dissemination of the software to the social science research community and its incorporation in graduate and postdoctoral education.