In search of reasonable doubt

A number of different approaches have been used to quantify jurors' use of the decision criterion known as reasonable doubt. The purpose of the present study was to determine which of a select group of these approaches could be utilized to match the actual decisions of individuals role-playing jurors for an assault trial. Simon's (1970) rank-order approach, a self-report approach, an approach derived from Statistical Decision Theory (Fried, Kaplan, & Klein, 1975), and one derived from Justice White's explanation of theJohnson v. Louisiana (1972) decision were each used to estimate values for reasonable doubt. The estimates were then used to recreate the individual decisions, and the recreated decisions were then compared to the actual decisions. The results indicated that every approach accurately matched the actual decisions at a better-than-chance rate, although the rank-order and decision theory approaches were most accurate. The reasonable doubt estimates obtained from each of the approaches were also used to examine the basic assumption underlying the Thomas and Hogue (1976) juror decision model. Only the estimates from the decision theory approach provided consistent support for the assumption. A variety of methods were suggested for more definite determinations of the accuracy of the approaches tested, and the relative merits of the approaches were discussed.