NCJ Number
217213
Journal
Journal of Forensic Sciences Volume: 52 Issue: 1 Dated: January 2007 Pages: 31-39
Date Published
January 2007
Length
9 pages
Annotation
This study examined whether using 16 or more markers (loci) increased the reliability of DNA-comparison evidence in criminal cases.
Abstract
Simulated results supported the hypothesis that analytical added value did not increase with the number of markers. The study showed that in a scenario that involved full siblings as a relevant population (the more extreme situation in favor of the defense), from the ninth marker on, the contribution of any new marker was extremely low (less than 0.05 percent). This means that increasing the number of markers beyond nine will add very little to the existing information about DNA characteristics. Therefore, it is not cost-effective to type more than nine DNA markers as a means of increasing the reliability of a DNA match. This paper proposes an approach that allows a forensic scientist to think systematically about a problem, while offering more structure and guidance in answering such questions as "What conclusions can be drawn from the available evidence?" and "Why is the chosen action appropriate?" The paper approaches these issues by using an example that involves DNA evidence in forensic science to justify the number of markers to type in a criminal case. Graphical models, such as influence diagrams, also called Bayesian decision networks, are introduced to deal with the scenario presented, i.e., whether more DNA markers must be typed in order to justify a DNA match between evidence found at a crime scene and the DNA of the defendant. 9 tables, 8 figures, and 23 references