As the use of statistics in litigation has burgeoned and as more complicated statistical techniques have entered the courtroom, concern for the way courts use statistics has mounted and efforts to instruct lawyers and judges on the wise use of statistics have begun. Professor Rubinfeld's paper is a contribution toward this end. Two ideas at the core of this paper are particularly important if we are to develop a more satisfactory approach to the use of statistics in the courtroom. The first is Professor Rubinfeld's caution against the talismanic use of the .05 level of significances as a test of what aspects of a statistical study are important to a legal factfinder. The second is his call for more attention to sensitivity testing than is customary in litigation research.
Lempert, Richard O. "Statistics in the Courtroom: Building on Rubinfeld." Columbia Law Review 85 (1985): 1098-1116.