Assessing the External Validity of Jury Research

点击翻译按钮获取中文摘要

9311660 PENROD Social scientists have studied jury decision-making in order to identify factors in the process that conform to the ideal model of the jury in a democratic society, as well as factors that might be changed to improve jury functioning. However, access to real juries is restricted by law and as a result many jury researchers use simulated juries as the basis for their studies. Simulated juries have the advantage that researchers can carefully control and systematically vary the content of their simulation to maintain experimental research controls not typically available in real courtrooms. Critics within the social science and legal communities argue, however, that the decision-making observed in jury simulations may not generalize to actual trials and that it may be unwise to base policies on the results of simulation research. This project will test whether the research methods used (those features of jury studies that arguably make them more or less applicable to actual cases) are systematically related to the size and nature of the results. A number of variables (such as jury size, jury decision rules, jury comprehension of legal instructions, and pretrial publicity) that have been examined using both simulations and field observations of real juries will be examined. The project will include a number of stages. First, a search will be undertaken to identify all relevant studies. Second, existing reviews of the jury literature and theoretical articles on jury behavior will be reviewed to identify variables for coding. A coding scheme for all included characteristics will be developed and tested. Then all of the studies will be coded according to the coding scheme. These data will be analyzed to test the average effect sizes produced by the independent variables in the full data set and the average effect sizes for specific research domains within the full data set. The relationships among the various research methods and the magni tude of the effect sizes will then be tested in both the full data set and the specific research domains. The goal is to identify any methodological variables that moderate the influence of the independent variables and thereby limit the generalizability of results from simulation studies. ***