Since the passage of the Civil Rights Act of 1964, various forms of employment discrimination have been prohibited by law. The courts have enforced these laws vigorously, often requiring plaintiffs and defendants to present quantitative analyses of applicant and employee records as evidence. In fact, statistical analyses have become a regular feature of employment discrimination cases. Consequently, there has been an explosion of interest in the production of statistical assessments of discrimination, as more and more social scientists and statisticians have become involved as consultants and expert witnesses on both sides of discrimination cases. This paper discusses current practices in the production of statistical assessments of discrimination and suggests nonstatistical techniques that can be used to aid the interpretation of the results of statistical analyses. Statistical assessments (1) may contradict accumulated social scientific knowledge about decision making, in general, and discrimination, in particular, and (2) may also contradict the spirit of civil rights legislation concerning discrimination. These shortcomings motivate the introduction of a method of analysis that incorporates principles of Boolean algebra, an approach that allows holistic comparison of categories of similarly situated individuals.