Sampling and Statistics

Abstract

In this chapter, I address several issues related to statistical sampling and statistical analysis when used as evidence in wage and hour litigation. Sampling and statistical analysis are commonly used in a variety of legal contexts including antitrust, employment discrimination, toxic torts, and voting rights cases. Properly designed and executed statistical analysis is generally considered admissible in litigation under the Federal Rules of Evidence as most sampling and analysis methods meet the “scientific knowledge” requirement in Daubert v. Merrell Dow Pharmaceuticals, a case which guides the admissibility of expert evidence in litigation. This chapter is focused on aspects of sampling and statistical analysis that are most frequently applied to address wage and hour disputes. This typically includes descriptive statistics to summarize data collected from a sample or estimates of population characteristics (“parameters”) based on data from a sample. Disputes over sampling and statistical analysis often play a major role in a court’s decisions to certify a class or determine liability. Disputes tend to be unrelated to the accuracy of the calculations but rather the reliability of the underlying data, the representativeness of the sample from which the data were collected, or the proper interpretation of statistical analysis results. Topics covered included statistical sampling, non-response bias, extrapolation and confidence intervals, variability, data quality, and damage calculations.