The Relationship between Disparate Impact Case Law and Empirical Research in the Behavioral Sciences

11 Pages Posted: 9 Jul 2015

See all articles by Palmer Morrel-Samuels

Palmer Morrel-Samuels

Employee Motivation and Performance Assessment; University of Michigan at Ann Arbor - School of Public Health

Date Written: July 8, 2015


Disparate impact litigation is unique, in part because outcomes often depend on statistical analysis from expert witnesses who are research psychologists. Accordingly, it is surprising that psychologists serving the court as expert witnesses in disparate impact (DI) cases often take one of two divergent roads, either focusing on case law, as Gutman et al. do in their helpful book, or on published peer-reviewed research in the behavioral sciences as Outtz does with similar effectiveness in his edited volume. Although there are prominent writers in both law and psychology who do address DI from an interdisciplinary perspective, knowledge transfer between experimental psychology and law is surprisingly meagre, even in exemplary texts such as Rutherglen and Donohue’s recent book Employment Discrimination; Law and Theory, (2012) which is comprehensive in virtually every other respect. This paper seeks to expand that important interdisciplinary exchange by examining the concordance — or lack thereof — between empirical findings in the behavioral sciences and important DI decisions. To compile an informative set of reliable empirical studies I followed (and adapted) conventional procedures for a meta-analysis by surveying empirical studies published in peer-reviewed journals, and by applying a filter to exclude studies where statistical significance, research methodology, or effect size fell below the standards articulated both in Rosenthal’s original work on meta-analysis and in the American Psychological Association’s article on standards for peer reviewed research. In the interests of enhancing the quality of expert reports and the legal decisions that rely upon them, this paper begins to build an interdisciplinary cross-reference linking DI case law and published empirical research in the behavioral sciences. I use two classic texts — Rosenthal and Rosnow’s Essentials of Behavioral Research: Methods and Data Analysis (2007), and Abelson’s Statistics as Principled Argument (1995) — to provide a theoretical foundation for evaluating empirical research so that this interdisciplinary effort can proceed judiciously. I cover two issues in depth: 1) Data Aggregation and the Omnibus Null Test; and 2) Multiple Regression and the Use of Covariates. I’ll close by briefly addressing causation — a pressing issue that deserves future consideration in a similar interdisciplinary analysis that bridges law and empirical research in experimental psychology.

Keywords: employment discrimination, disparate impact, adverse impact, empirical research, statistics, scientific evidence, data aggregation, multiple regression, omnibus null, covariates

JEL Classification: C00, C10, C12, C43, J70, J71

Suggested Citation

Morrel-Samuels, Palmer and Morrel-Samuels, Palmer, The Relationship between Disparate Impact Case Law and Empirical Research in the Behavioral Sciences (July 8, 2015). Available at SSRN: or

Palmer Morrel-Samuels (Contact Author)

University of Michigan at Ann Arbor - School of Public Health ( email )

1415 Washington Heights
Ann Arbor, MI 48109-2029
United States
734-368-3348 (Phone)

Employee Motivation and Performance Assessment ( email )

210 Park St.
Chelsea, MI 48118
United States
734-368-3348 (Phone)


Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Abstract Views
PlumX Metrics