Citation

Vergeer P, Hendrikse JN, Grutters MMP, Peschier LJC. Sci. Justice 2020; 60(5): 438-450.

Copyright

(Copyright © 2020, Forensic Science Society, Publisher Elsevier Publishing)

DOI

10.1016/j.scijus.2020.06.002

PMID

unavailable

Abstract

In many arson cases gasoline is used as a means to start the fire. In this paper results are presented for a likelihood-ratio (LR) system aimed at comparing gasoline traces from fire debris to a reference gasoline. The LR-system is able to deal with disturbing effects caused by burning and exposure to surroundings: pyrolysis products, preferential adsorption and evaporation. This paper focusses on the criminalistic and statistical aspects of the design of the LR-system, and presents results on performance of the LR-system. The details of trace gasoline recovery from fire debris will be presented in an accompanying paper. Validation and performance measures show that this system gives well-calibrated LRs for comparisons involving trace samples with a spread in quantity of gasoline, evaporation levels, and matrices that are typically encountered in casework. Rates of misleading evidence are less than 3.5%. We conclude that, despite limitations in experimental design, this LR-system can be useful to the comparison of gasoline profiles in casework practice.

Language: en