NCJ Number
179130
Journal
Law and Human Behavior Volume: 22 Issue: 4 Dated: August 1998 Pages: 389-407
Date Published
August 1998
Length
19 pages
Annotation
Findings of experimental studies and an analysis of police records from 115 cases in Israel in 1992 were used to determine the extent to which judges and juries overbelieve eyewitnesses who choose suspects in lineups.
Abstract
The extent of the problem is known to depend on the probability that the defendants who were chosen in lineups were actually guilty. Bayes' theorem notes that the probability of their guilt depends as much on the relative number of guilty who are chosen as on the number of innocent suspects. However, both the experimental data and the analysis of archival reports of real eyewitness cases suggest probabilities that appear to be conservative estimates. These findings lead to the probability of 0.247 of being innocent if chosen in a lineup, assuming no a priori assumption of guilt or innocence. Thus, the problem of overbelieving is serious. Options for addressing this problem include ignoring it, banning lineups, limiting lineup testimony to cases where additional evidence is available, educating the judge and jury regarding the appropriate value of the lineup, and using the expert witness to differentiate accurate and inaccurate eyewitnesses by sensitizing the jury to different eyewitness conditions. Tables, appended list of experimental studies used in the analysis, and 64 references (Author abstract modified)