Kerr, N. L., & MacCoun, R. J. (1985). The effects of jury size and polling method on the process and product of jury deliberation. Journal of Personality and Social Psychology, 48, 349-363.
The U.S. Supreme Court has repeatedly assumed the functional equivalence of different sized juries (at least in the range of 6- to 12-person groups). Several formal models of jury decision making predict that larger juries should hang more often, particularly for very close cases. Failures to confirm this prediction in several previous studies were attributed to inadequate sample sizes or to insufficiently close cases. An experimental simulation study that minimized these problems was undertaken to test the models’ prediction. Social decision scheme and social transition scheme analyses permitted comparisons of the decision-making processes of the different-sized mock juries. The effect of the method used to poll group members’ verdict preferences was also examined. As group size increased, the observed probability of a hung jury increased significantly. No process differences between 6- and 12-person groups were detected, but 3-person groups did exhibit several process differences from the larger groups. When cases were very close, the likelihood of a hung jury for typically sized juries was found to be lower when the group was polled by secret ballot than when a show-of-hands polling method was used.
Larger juries (6 and 12) more likely to hang and took longer to deliberate; no impact of jury size on convictions. No effect of polling secrecy on hung-jury rate when cases were clear, but smaller juries hung less in close cases with public polling whereas larger juries hung more. SDS, social transition schemes, and transition speed were similar for 6- and 12-person juries. Preference shifts were a function of relative and absolute faction size.