NCJ Number
172961
Date Published
1998
Length
2 pages
Annotation
This research abstract describes a study that used several forensic databases compiled by the Federal Bureau of Investigation (FBI) and determined the numerical effects of a population structure on the weight to be attached to matching DNA profiles from a suspect and material from a crime scene.
Abstract
When the DNA profile from a suspect matches the DNA profile from crime-scene material, it is usual to estimate the frequency of that profile in the population. However, these estimates have been criticized on the grounds that they do not account for the population substructure. Presenting the general population frequency will overstate the match if the suspect and perpetrator belong to a particular group of people in which the matching DNA profile is more common than in the general population. This criticism can be met by expressing the strength of the evidence as a likelihood ratio that explicitly accounts for population structure by means of a parameter known as FST. This parameter was estimated from several forensic databases compiled by the FBI to quantify the effects of population substructure on the estimated frequencies of matching DNA profiles. The analysis revealed that the effects of population are numerically small in large populations such as white people and black people, unless the allele frequencies are smaller than those generally encountered.