Wharton Research Scholars

Document Type

Journal Article

Date of this Version



Simpson’s paradox refers to the reversal of a statistical relationship between two variables in sub-populations when the sub-populations are combined and analyzed as a population. This article is intended to provide a broad survey of the past, present and future research surrounding the issue. Real data from a discrimination litigation case is examined to identify the occurrence of the paradox.



Date Posted: 29 September 2006

This document has been peer reviewed.