NCJ Number
218789
Date Published
January 2001
Length
3 pages
Annotation
This article critiques the statistical methods used by two recent authors in this journal to interpret DNA profiles.
Abstract
The author argues that in both cases of statistical method error, the errors could have been avoided by making distinctions between one-person profile probabilities and two-person match probabilities. It is statistically important to distinguish between “profile probability” when describing the chance of a single individual having a particular profile and “match probability” when describing the chance of an individual having the profile when it is known that another individual has the profile. However, the distinction between profile and match probabilities are rarely made by practicing forensic scientists. Yet, the author points out that profile probabilities, while having some forensic value, are unlikely to be relevant in forensic calculations. The author calls on forensic scientists to advance the science by applying correct equations and calculations in their data analyses. The particular errors made by Budowle et al. (2000) and Fung and Hu (2000) are described and the author shows how the analysis should have been conducted. References