NCJ Number
198550
Journal
Law and Human Behavior Volume: 26 Issue: 6 Dated: December 2002 Pages: 625-639
Date Published
December 2002
Length
15 pages
Annotation
This article examines what is expected in jury decision making and what actually occurs.
Abstract
Jurors make two decisions: determining culpability of the defendant and how much evidence is required to convict. Jurors need three theoretical tools to make these decisions. The first is expected utility theory, which implies that the juror will acquit if the expected utility of acquitting exceeds that of convicting. The next tool is Bayes’ Theorem, which analyzes the odds of guilt. The third tool is called signal detection theory, which says that a juror that wants to maximize the expected utility of a decision should listen to the evidence that is sampled from an unknown distribution. The expected utility theory, Bayes’ Theorem, and signal detection theory can be used to show that the diagnostic performance of jurors is extremely likely to be far lower than expectation, irrespective of the jurors’ willingness to base their decision on probabilistic considerations or any other cognitive strategy. Public opinion surveys indicate that most people want errors of convicting the innocent or acquitting the guilty to fall well below 10 percent. Through use of these three theories, it is argued that the frequency of mistakes probably far exceeds these “tolerable” levels. Two potential solutions to this problem are to adjust for prior beliefs of jurors and to have jury decisions made by groups rather than individuals. Upon further analysis, these two solutions are not viable. Even though jury decisions are highly regarded methods for determining justice, it must be understood that juries make more mistakes than most would think. 2 figures, 2 tables, 6 footnotes, appendix, 49 references