A Model of the Jury Decision Process

A computer simulation of jury decision making.

“…Changes of the individual juror's opinion are modeled as movement from verdict state to verdict state, and this movement is a probabilistic function. To model this process we have treated deliberation and persuasion as a discrete-time Markov process and developed a model (DICE) related to models suggested by Davis (1980), Godwin and Restle (1974), Klevorick and Rothschild (1979), and Stasser and Davis (1977). In DICE we assess the voting status of each juror and the size of factions within the jury at fixed time intervals, or ballots.…”

The optimal jury size when jury deliberation follows a random walk

“…If k out of N jurors support conviction on the initial ballot, then the probability that the jury eventually votes to convict the defendant, Pr(Convict), is k/N . If the jury unanimously supports one outcome on the initial ballot, there will be no deliberation and this will be the trial verdict (Hastie et al 1983;Klevorick and Rothschild 1979).…”

“…If, for example, instead of assumption 2, the transition probabilities in the Markov chain were to be dependent on the current stage (as in Klevorick and Rothschild 1979) it can be demonstrated that the probability of a type I error is always less than 1 − p as long as N is greater than 2. Furthermore, under this alternative assumption the probability of both types of errors decreases as N increases (approaching 0 as N become "large" and p > 0.5).…”

Chameleon voters and public choice

“…Moreover, almost all juries succeed in achieving a consensus; only about 5 percent are 'hung,' or divided. 24 The hung jury is treasured by the legal community as a sign that members with minority viewpoints are able to resist majority pressure. At the same time, the community feels that too many hung juries would impede the court system's effectiveness.…”