Development and Implementation of a Fire Debris Inference Validation Database: Moving to Probabilistic Statements of the Strength of the Evidence

As submitted by the proposer:

The goals of this research are to develop a fire debris inference validation database in collaboration with the fire debris analysis community, and use the database to develop and test both rules-based methods and numerical methods to enable probabilistic assessment and reporting of the weight of the evidence.

The proposed research will be conducted in collaboration with members of the fire debris community as an ad hoc committee of the Technical and Scientific Working Group for Fire and Explosions (T/SWGFEX). Achieving the goal of probabilistic reporting requires access to a set of samples that are representative of casework and for which the ground truth is known. In other words, each sample is known to be either positive or negative for ignitable liquid residue, ILR. An open-access database of complex samples comprised of mixtures of partially evaporated ignitable liquid residue (ILR) and substrate pyrolysis products does not currently exist.

The goals of the research will be met through a set of specific objectives: Objective 1. In the first six months, development of a laboratory method for creating realistic fire debris analytical samples that are representative of casework samples.

Objective 2. Over the course of two years, populate an on-line database of 800 samples generated following the methods developed under objective 1.

Objective 3. While the database is being populated, we will work with the T/SWGFEX committee to produce a decision tree and subsequent rules-based protocol following current laboratory practices for assessing the presence of ILR in a fire debris sample. The rules-based protocol and the database records will be used to assess the probability of observing the evidence when a fire debris sample contains ILR, as opposed to when it does not contain ILR.

Objective 4. In the third year of the project, the researchers will test and validate the rules-based method developed in objective 3, and previously reported numerical methods using the database developed in objective 2.

Objective 5. Throughout the project, the researchers will disseminate the results of the research through open access to the database, publication of the results and training of fire debris analysts.

The results from this work will influence forensic research and practice by demonstrating how methods based on expert knowledge can be implemented in such a way as to facilitate the move from categorical statements to probabilistic statements of evidentiary value.

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).

ca/ncf