A Jury Experiment Reanalyzed
Researchers in the behavioral sciences have watched with some pride as the courts have given increased attention to social science studies. Judicial interest in empirical studies is a desirable development but one not quite free of danger. The courts are not yet fully accustomed to dealing critically with such evidence. The United States Supreme Court ruled recently, in Colgrove v. Battin, that six-member juries in civil cases meet the seventh amendment requirement of trial by jury. This decision was not surprising in light of Williams v. Florida, in which the Court ruled that six jurors were sufficient to satisfy the sixth amendment jury trial guarantee in a criminal trial in state court. In both decisions, the Court claimed to be convinced that there would be no difference in trial outcomes if the size of the jury were cut in half. In the Colgrove decision, four empirical studies were cited as “convincing empirical evidence” in support of this position; two of these studies had been published in the University of Michigan Journal of Law Reform. It has been shown elsewhere, in summary fashion, why none of the four studies supplied valid information concerning the issue of whether jury size affects trial outcome. This article considers one of these studies in detail, because it provides a good example of the inappropriate use of social science data. Conclusions are presented which may mislead the unwary reader and important information has been omitted, however unintentionally, from the study report. Because these data were missing, all of the tables in this review had to be developed from materials not presented to readers of the original report. This critique is designed primarily to help future researchers avoid repetition of these errors.