U.S. flag

An official website of the United States government, Department of Justice.

Statistical Infrastructure for the Use of Error Rate Studies in the Interpretation of Forensic Evidence

Award Information

Award #
Funding First Awarded
Total funding (to date)

Description of original award (Fiscal Year 2020, $87,206)

The forensic identification of source problem is a fundamental question of interest in forensic science and is concerned with whether or not a set of traces with an unknown origin arose from a specified source of traces. This question forms the basis upon which more complicated questions, such as activity-level questions, are built. This question is usually restated in the context of two competing propositions, one associated with the prosecution model that a specified source is the source of the traces and one associated with the defense model that a source in some relevant background population is the actual source of the traces. This restatement of the question reframes the forensic identification of source problem as a nonnested model selection problem. Recently, the National Institute of Standards and Technology has been tasked with organizing and supporting the broad community of forensic scientists with the goal of developing general standards for the implementation and practice of forensic science. One of the more contentious efforts concerns the standards for expressing conclusions in a forensic identification of source problem. In part, this is due to several distinct statistical traditions being used to quantify uncertainty in forensic source conclusions. In this research program we will focus on the use error rate based methods to solving the forensic identification of source problem, with the specific goal of building up a suite of statistical tools to support long term research into the interpretation and presentation of the value of forensic evidence using error rates. It is expected that this work will lead to at least 4 different research programs that will be appropriate to be extended in PhD dissertations in mathematical statistics and statistical pattern recognition. Manuscripts on findings will be promptly prepared and submitted for publication. The computer codes and the user guides will be made publicly available on investigators' websites. CA/NCF Note: This project contains a research and/or development component, as defined in applicable law," and complies with Part 200 Uniform Requirements - 2 CFR 200.210(a)(14).

Date Created: September 21, 2020