A data analysis of success in US Officer Candidate School, the use of ASVAB waivers, and race

The paper takes an in-depth look at the controversy posed by the facts that the use of ASVAB score waivers for admission to Officer Candidate School (OCS) appears: (1) unrelated to success in OCS when viewed by the individual races; and (2) related to success in OCS when the data are pooled into a single macro set. The short answer is found in the fact that the use of waivers is quite variable from race to race. Further, increasing use of the waiver is associated with decreasing success rates in OCS. It is noted that the use of waivers diminished during the period of the study. The general result would be spurious if the OCS has some sort of racial bias internal to it. Another explanation is that the ASVAB or the administration of the decision rules has a bias that accepts candidates by race group, leading to uneven success rates in the school.