A Differential Item Functioning Analysis of the Alternative Model of Personality Disorders Across Younger and Older Age Groups.

Abstract

Age bias on personality pathology measures creates challenges regarding assessment of personality disorders among older adults. This study cross-sectionally examined age bias on measures of the Alternative Model of Personality Disorders' (AMPD) two diagnostic criteria. Younger and older adults in the United States completed the Levels of Personality Functioning Scale-Self-Report (LPFS-SR) and Personality Inventory for Diagnostic and Statistical Manual of Mental Disorders, 5th Edition-Brief Form (PID-5-BF). Differential item functioning (DIF) analyses examined bias on the item-level whereas differential test functioning (DTF) analyses investigated impacts on the domain-level. For the LPFS-SR, 23% of items demonstrated DIF, with the Empathy and Intimacy domains most impacted. For the PID-5-BF, 40% of items showed large DIF, with four of five domains demonstrating large age bias. Across measures, items showed evidence of bias for and against older adults. Findings indicate meaningful age bias on measurements of both AMPD diagnostic criteria, suggesting greater potential to mischaracterize findings when using these measures across the lifespan.