U.S. flag

An official website of the United States government, Department of Justice.

NCJRS Virtual Library

The Virtual Library houses over 235,000 criminal justice resources, including all known OJP works.
Click here to search the NCJRS Virtual Library

Meta-Analytically Quantifying the Reliability and Biasability of Forensic Experts

NCJ Number
223880
Journal
Journal of Forensic Sciences Volume: 53 Issue: 4 Dated: July 2008 Pages: 900-903
Author(s)
Itiel Dror Ph.D.; Robert Rosenthal Ph.D.
Date Published
July 2008
Length
4 pages
Annotation
This study used meta-analytic procedures and estimate-effect sizes in indexing the degree of reliability and bias of forensic experts, based on within-expert comparisons in which the same expert unknowingly made judgments on the same data at different times.
Abstract
The findings show that forensic experts (fingerprint experts in the current study) are not totally reliable nor are they unbiased. Regarding reliability, 89 percent of the experts reached the same conclusions on fingerprint matches as they had made in their initial analysis (obtained from archives without informing participants they had made previous decisions in these cases). This is significantly better than chance but is far from what is the accepted and recognized norm for fingerprint expert performance. In testing bias, extraneous information was added, so as to cause the experts to have expectations about the outcome of the examination. Although these findings are based on fingerprint experts’ decisionmaking, because this domain is so well established, they apply equally well, if not more, to all other less established forensic domains. The study involved six fingerprint experts, each with more than 5 years experience in latent fingerprint examination, post training, and accreditation. They all consistently passed all their proficiency testing and were regarded as very competent examiners by their supervisors. Their past decisions on fingerprint analyses were retrieved from archives of real criminal cases and then represented to the same experts. The fingerprints were given to the experts in the same format that they were presented to them in the past in order to determine reliability. In order to determine whether bias was involved in decisionmaking, extraneous information was added for some of the cases. 3 tables and 11 references