NCJ Number
248829
Date Published
August 2015
Length
40 pages
Annotation
This study examined the effects of two techniques for quantifying and reducing bias in expert testimony, so as to reduce one of the underlying causes of wrongful convictions.
Abstract
One of the techniques examined is the "blinding" of expert witnesses, which means that prior to their testimony they do not know whether they will be testifying for the prosecution or the defense. The second technique provides consensus feedback on an expert's findings prior to his/her testimony. The study found that a consensus of experts regarding an expert's findings prior to his/her testimony resulted in fewer scientifically invalid conclusions. The study determined that without a consensus of qualified experts regarding a single expert witness's findings, 88 percent of the findings were incorrect. Thus, expert consensus feedback decreased the rate of erroneous expert testimony. On the other hand, the use of the "blinding" technique did not result in fewer errors in expert testimony. The hypothetical case developed for the study involved criminal trespass and grand theft larceny of property belonging to a private business. The forensic question to be answered by the expert witness was the probability that the defendant used the stairwell. The solution could be derived by using Bayesian probability techniques familiar to those with experience or academic training in a variety of physical and social science fields such as math, chemistry, psychology, criminology, and law. In the consensus exercise, a panel of 12 scientists with doctorates in relevant fields evaluated the case in order to reach a consensus about the best answer to the probability question. 5 figures, 4 tables, and a 61-item bibliography
Date Published: August 1, 2015