Personality Assessment Inventory (PAI): obsolete norms identify psychopathology in nearly everyone

The Personality Assessment Inventory (PAI), a self-report personality test, is one of the frequently used measures to assess psychopathology in a wide variety of settings including in high stakes assessments, for example, in child custody disputes, employment settings and fitness for duty assessments. The PAI has never been normed on a Canadian population and the PAI users have simply assumed that the US norms also describe the Canadian population. Moreover, accumulated research demonstrates that the PAI's 35 years outdated and obsolete norms no longer describe neither university students' nor normal adult US populations.

We administered the PAI to over 200 university students in a mid-size Canadian university.

Our students scored on average in moderately elevated range (60T to 69T) on many of the PAI scales including anxiety (ANX), anxiety-related disorders (ARD), depression (DEP), schizophrenia (SCZ), and borderline features (BOR). Multivariate base rate analyses revealed that approximately 95% of our sample scored in elevated range on at least one out of the 22 PAI Scales. Furthermore, although some of the PAI reliabilities are adequate for research, the PAI reliabilities are too low for using the PAI in high stakes and forensic assessment, for example, in insurance benefits, child custody, employment, and fitness for duty assessments.

We conclude that the PAI US norms are no longer appropriate for high-stakes assessments, ought to be withdrawn immediately, and new up-to-date norms ought to be established to prevent mislabelling and diagnostic misclassifications of and harm to examinees. Continued use of the PAI outdated norms in high stakes assessments carries ethical risks, is non-scientific, and likely amounts to malpractice.