Data analysis in forensic science: A Bayesian decision perspective

The use of formal statistical methods to analyse quantitative data in forensic science has increased considerably over the last few years. Students, researchers and practitioners in forensic science regularly ask questions concerning the relative merits of different approaches, in particular the frequentist and Bayesian approaches, to statistical inference in the forensic context. The ideas of the Bayesian approach in forensic science are now being extended to include decision theory and the associated concept of utility. The book sets forth procedures for data analysis that rely on the decision-theoretic approach to inference. Emphasis is made on foundational philosophical tenets as well as the implications of the decision-theoretic approach in practice, and a range of statistical decision-theoretic methods that are useful in the analysis of forensic scientific data is discussed. Forensic scientific examples include point estimation, the comparison of means and proportions in populations, the chioce of sample size and the classification of items of evidence of unknown origin into predefined populations. Graphical models (e. g. Bayesian networks) are used to illustrate selected applications of Bayesian methodology.