The Validation Trap: Why Personality Tests Remain a Hiring Liability
Personality tests may predict job performance yet still discriminate against disabled applicants, exposing companies to legal risk despite professional development.
Target Corporation's $2.8 million settlement over a pre-employment personality test exposed a critical gap: validation and fairness are not synonymous. While standard validation studies confirm tests predict job performance, they rarely analyze pass rates by disability status. The EEOC argued such tests can function as medical examinations under the ADA, and companies cannot rely on vendor certificates of validity as protection against adverse impact claims. The key question is whether organizations can produce their own data demonstrating that their specific use of personality assessments does not disproportionately screen out qualified applicants with disabilities.
In 2015, Target Corporation paid $2.8 million to settle a disability discrimination suit over its pre‑employment personality test. The assessment was not a fringe internet quiz. It was a professionally developed product from a respected industrial‑organizational psychology firm, and it had passed standard validation checks. The settlement still reverberates because it exposed a gap most companies do not see: validation and fairness are not the same thing. The test measured traits like sociability, emotional stability, and stress tolerance. Candidates were asked to agree or disagree with statements such as “I often feel nervous in social situations” and “I prefer to work alone rather than in a group.” Those items look neutral on a psychometric spec sheet. For an applicant with an anxiety disorder, depression, or autism, they measure more than a preference—they probe the edges of a disability. The Equal Employment Opportunity Commission argued the test functioned as a medical examination under the Americans with Disabilities Act, and that Target had not shown it was job‑related and consistent with business necessity. Standard validation studies typically check whether a test predicts job performance or whether its factor structure holds across large samples. They rarely break down pass rates by disability status. The test’s developers had done the first kind of work. They had not demonstrated that the test did not disproportionately screen out people with the very conditions the items touched on. The professional validation a vendor ships with the product is usually a general‑population study. It does not tell you what happens when the tool hits your actual applicant pool, where patterns of disability, race, and gender combine in ways no manual can anticipate. Target settled without admitting liability, but the case put the industry on notice. A vendor’s certificate of validity is not a shield against an adverse impact claim. If you cannot show, with your own data, that your personality test does not disproportionately knock out qualified applicants with disabilities, you are carrying a risk the contract does not cover. The question is not whether the test is well‑built. It is whether you can produce the evidence that your specific use of it is fair.