Measurement invariance of the Personality Inventory for DSM-5 across sex



Rivard G, Le Corff Y, Lapalme M, Forget K. (2023). Measurement invariance of the Personality Inventory for DSM-5 across sex. Journal of Personality Assessment. 15, 1328937.


Introduction : There has been an international movement towards dimensional models of personality disorders (PDs) in the last decades, which culminated in the publication of the Alternative Model of Personality Disorders (AMPD) in the Emerging Measures and Models section of the DSM-5. This model was accompanied by a APA-sanctioned Personality Inventory for DSM-5 (PID-5) for the assessment of the AMPD pathological personality traits. One major issue with the assessment of personality disorders pertains to sex differences, and measurement invariance across sex in assessment instruments for PDs is necessary in order to ensure non-biased evaluations and to make valid comparisons between men and women. This study aimed to provide more information on measurement invariance across sex for the PID-5, using both the original scoring approach provided by the authors of the instrument and the scoring approach suggested by the APA in the published version of the PID-5.

Methods : This study was conducted with a sample of 2273 participants from the general Québec (Canada) adult population aged 18 to 90 years (M = 46.59; SD = 16.32; 51.8% women).

Results : The original scoring approach model showed good fit to data after freeing paths between certain traits and reached strict invariance. The APA scoring approach also showed good fit to data and reached strict invariance, but needed an adjustment (path freed between Emotional lability and Impulsivity in men) to reach scalar invariance.

Discussion : In line with previous research, the PID-5 is invariant across sex and the five-factor structure adjusts well to data. The APA scoring approach appears to attenuate the cross-loading problem observed with the original scoring approach. In light of these results, we recommend using the APA scoring approach to derive domain scores.