Uniquely Biased

How ASWB Exams Violate Psychometric Best Practices

Abstract

After publication of the 2022 Pass Rate Analysis demonstrating that minoritized social workers pass at less than half the rate of white social workers, the Association of Social Work Board (ASWB) Examination Guidebook (ASWB, 2023a), revised its psychometric reporting of exam fairness from “statistically free from race and gender bias” (ASWB, 2022a, p. 3) to “differences in exam performance for…different demographic groups…is influenced by many factors external to the exam,” upstream of the examination in the workforce pipeline (ASWB, 2023a, p. 9). Focusing only on factors external to the exam ignores the possibility that the internal properties of the exam may be invalid, unreliable, and unfair. Race, class, culture, and other structural factors have not impacted ASWB exams the same over time, with ASWB’s 2022 Exam Pass Rate Analysis reporting 10-13% reductions in bachelor’s and master’s examination pass rates after introducing the 2018 exam blueprint. Using extensive references to ASWB’s public statements, this article will demonstrate how ASWB elided evidence of examination flaws and presented external factors as the only possible explanation for disparities in pass rates. Beginning with the policy paradox created by national organizational disagreement on the cause and next steps on exam score inequities, this article will demonstrate how bias is encoded in the language and theories underlying the examination, as well as review the extant empirical evidence on psychometrics addressing shortcomings in ASWB’s exam validation process, which converge to create a uniquely biased exam.