The appeal of Bayesian statistics is its intuitive basis for making direct probability statements for all assertions including cases that involve disparate data types. In scientific disciplines, where it is often necessary to take existing knowledge and update it in light of new data, Bayesian models are singularly powerful. Until recently, the cost of computation, the knowledge barrier, and the absence of commonly accessible computational tools have been barriers. Powerful computation is now more widely accessible, and a generation of powerful, open source, free, and easier to use software for Bayesian statistics are now available. Additionally, work of Bayesian experts over the years has produced user-friendly conceptual explanations of the Bayesian idea that can help every practitioner of applied science with interest in using the Bayesian approach. This talk assumes no prior experience with Bayesian statistical modeling and is intended as both a non-mathematical introduction to the theory as well as a practical introduction to the use of readily available and easy-to-use tools. An understanding of Bayesian statistical modeling will be developed by relating it to existing knowledge of traditional approaches. Some of the underpinnings and departures from conventional frequentist interpretations of probability will be explained in order to motivate the development of Bayesian statistical modeling. A step by step walk through of examples for some of most common and familiar statistical tasks with instructions and code for implementation is presented. The examples will focus on the core of most common statistical tasks related to small molecules, with code made available on GitHub for easy access and experimentation.
