Adverse Impact Is Far More Complicated Than the Uniform Guidelines Indicate

Abstract

McDaniel, Kepes, and Banks (2011) pointed out many critical flaws that were either contained in the Uniform Guidelines on Employee Selection Procedures

Encouraging Debate on the Uniform Guidelines and the Disparate Impact Theory of Discrimination

“…Consistent with the position of the focal article, Barrett, Miguel, and Doverspike (2011) asserted that we can expect such differences for at least the next 100 years. Jacobs, Deckert, and Silva (2011) noted that “eliminating group differences is not possible for many effective predictors.” We offer that it is best to recognize that mean racial differences in job‐related attributes exist, do not seem to be going away, and need to be addressed.…”

“…There are two themes in the commentaries concerning adverse impact. One relates to the measurement of adverse impact (e.g., Barrett et al, 2011; Jacobs et al, 2011; Mead & Morris, 2011) and the other, perhaps more importantly, concerns the disparate impact theory of discrimination as a more central problem for personnel selection than the Uniform Guidelines (e.g., Barrett et al, 2011; Dunleavy, Aamodt, Morgan, Gutman, & Cohen, 2011; Jacobs et al, 2011; Sharf, 2011). Each theme will be addressed in turn.…”

“…Concerning the measurement of adverse impact, Jacobs et al (2011), Mead and Morris (2011), and Barrett et al (2011) addressed problems with statistical tests to assess adverse impact. It was argued that the 4/5th rule, which is explicitly addressed in the Uniform Guidelines , or statistical significance tests, which are used more frequently by federal agencies and the courts, have shortcomings that make them nonoptimal to assess the presence of adverse impact accurately (Barrett et al, 2011; Jacobs et al, 2011; Mead & Morris, 2011; Morris & Lobsenz, 2000; Roth, Bobko, & Switzer, 2006). Moreover, different methods often produce differing results, and at least one of them usually indicates the presence of adverse impact (Barrett, Doverspike, & Young, 2010; Barrett et al, 2011; Dunleavy & Gutman, 2009).…”

“…Then, such experts can claim that the selection test is not valid or that a better alternative test exists (Barrett et al, 2011). Also, Sackett (2011) and Jacobs et al (2011) raised important concerns over expert testimony. We agree with Barrett et al's (2011) assertion that even the elimination of the 4/5th rule from a revised Uniform Guidelines would not solve the problem.…”

Hispanic Representation in the Department of Defense Civilian Workforce: Trend and Barrier Analysis

“…Lack of familiarity with and understanding of statistical concepts, including the influence of sample sizes, creates confusion and hinders decisionmakers' ability to assess whether large discrepancies are present. There have been instances within the adverse-impact context in which use of the 80-percent and the two-standard deviation rule on the same data yield different conclusions on whether meaningful discrepancies are present (Jacobs, Deckert, and Silva, 2011;Peresie, 2009). …”