The Characterization of Monte Carlo Errors for the Quantification of the Value of Forensic Evidence

• NCJ Number: 252364
• Journal: Journal of Statistical Computation and Simulation Volume: 87 Issue: 8 Dated: 2017 Pages: 1608-1643
• Author(s): Danica M. Ommen; Christopher P. Saunders; Cedric Neumann
• Date Published: 2017
• Length: 36 pages
• Grant Number(s): 2014-IJ-CX-K088
• Agencies: NIJ-Sponsored
• Publication Type: Research (Applied/Empirical)
• Award: Foundational Research into the Quantification of the Value of Forensic Evidence for Complex Evidential Forms arising from Impression and Pattern Evidence

Annotation

This article reports on the development of a generally applicable method for characterizing the numerical error associated with Monte Carlo integration techniques used in constructing the Bayes Factor.

Abstract

Recent developments in forensic science have precipitated a proliferation of methods for quantifying the probative value of evidence by constructing a Bayes Factor that allows a decision-maker to select between the prosecution and defense models. Unfortunately, the analytical form of a Bayes Factor is often computationally intractable. A typical approach in statistics uses Monte Carlo integration to numerically approximate the marginal likelihoods composing the Bayes Factor. The current article describes the derivation of an asymptotic Monte Carlo standard error (MCSE) for the Bayes Factor, and its applicability to quantifying the value of evidence is explored, using a simulation-based example involving a benchmark data set. The simulation also explores the effect of prior choice on the Bayes Factor approximations and corresponding MCSEs. (Publisher abstract modified)