Unobtrusive measurement of racial bias in job-placement decisions

Abstract

Unobtrusively gathered historical data documenting recruit-classification decisions made by the US Navy were utilized to determine whether racial bias exists among classifiers. The nature of the classification procedure resulted in the essentially random assignment of over 27,000 Black and White recruits to 46 Black and White classifiers. This permitted a number of interesting comparisons and obviated numerous problems inherent in racial-bias studies. The major hypothesis that Black and White classifiers would be differentially biased in their treatment of Black and White recruits was not supported. A 2nd hypothesis that classifiers within either racial group would be differentially biased in their treatment of Black and White recruits also was not confirmed. (PsycInfo Database Record (c) 2025 APA, all rights reserved)