Bayesian radiocarbon modelling for beginners

Abstract

Due to freely available, tailored software, Bayesian statistics is now the dominant paradigm for archaeological chronology construction in the UK and much of Europe and is increasing in popularity in the Americas. Such software provides users with powerful tools for Bayesian inference for chronological models with little need to undertake formal study of statistical modelling or computer programming. This runs the risk that it is reduced to the status of a black box, which is not sensible given the power and complexity of the modelling tools it implements. In this paper we seek to offer intuitive insight to ensure that readers from the archaeological research community who use Bayesian chronological modelling software will be better able to make well educated choices about the tools and techniques they adopt. Our hope is that they will then be both better informed about their own research designs and better prepared to offer constructively critical assessments of the modelling undertaken by others.