Despite a lively debate about the dimensional vs. categorical nature of Personality Disorders (PDs), direct empirical tests of the underlying structure are missing for most PDs. Taxometrics can be used to investigate whether latent structures are categorical or dimensional. We investigated the latent structure underlying Avoidant, Dependent, Obsessive-Compulsive, Depressive, Paranoid, and Borderline PD by means of three types of taxometric analyses. SCID-II based DSM-IV PD criterion scores from 1,816 patients from Mental Health and Forensic Institutes, and 63 nonpatients, were analyzed with three types of taxometric analyses. MAMBAC, MAXEIG, and L-MODE taxometric analyses were applied on multiple criteria sets, constituted both on theoretical grounds and randomly. Assumptions for taxometric analyses were generally met. All but two of the 78 taxometric analyses indicated greater evidence for a latent dimensional structure, with better fit of empirical data to dimensional than to taxonic simulations; mean Comparative Curve Fit Index (CCFI) =.23, SD = .09. Only two analyses yielded ambiguous evidence (CCFI in the .40-.60 range) and none indicated taxonic structure.
Arntz, A., Bernstein, D., Gielen, D., van Nieuwenhuyzen, M., Penders, K., Haslam, N., & Ruscio, J. (2009). Taxometric evidence for the dimensional structure of cluster-c, paranoid, and borderline personality disorders. Journal of Personality Disorders, 23(6), 606-628. https://doi.org/10.1521/pedi.2009.23.6.606