The model rests on decision theory and includes two types of juror beliefs: the juror's degree of belief that the accused person is actually guilty of the crime and the juror's personal standard of what constitutes reasonable doubt. Both types of beliefs influence juror decisionmaking, because the juror knows that in human fact-finding systems it is possible to falsely convict an innocent person or falsely acquit a guilty person. However, the numerical values that jurors assign to the concept of reasonable doubt vary significantly, depending on whether jurors are asked directly about the level of certainty they require for a guilty vote or whether they are asked to assign values to the four possible outcomes of their votes. These outcomes are a conviction of a guilty person, a conviction of an innocent person, an acquittal of a guilty person, and an acquittal of an innocent person. The disparity that emerges from the two ways of eliciting juror views is not the result of the methodology used to gather these views, however. Instead, it reflects individuals' lack of clarity about the basic issue involved: whether acquitting a guilty person or convicting an innocent person is a more serious problem. Figures and 25 references.