In this article from the Journal of Forensic Economics, NERA Senior Vice President Dr. David Tabak examines the effects of supporting a claim or affirmative defense via statistical analysis in cases where passing one of many tests will suffice. While this concept is referred to in the Federal Judicial Center’s Reference Manual on Scientific Evidence, Dr. Tabak takes the further step of tying it to the Daubert criterion of examining the known or potential error rate in an analysis. He illustrates how a failure to account for multiple tests will inflate the error rate of an analysis and uses examples of labor discrimination and securities fraud claims to illustrate when and how one can quantify and correct for the bias in the error rate.
This article, from the March 2007 issue of the Journal of Forensic Economics, has been reproduced with permission from the publisher. All rights reserved.