Development and Validation of Personality Disorder Spectra Scales for the MMPI–2–RF

ABSTRACT

The purpose of this study was to develop and validate a set of MMPI–2–RF (Ben-Porath & Tellegen, 2008/2011) personality disorder (PD) spectra scales. These scales could serve the purpose of assisting with DSM–5 PD diagnosis and help link categorical and dimensional conceptions of personality pathology within the MMPI–2–RF. We developed and provided initial validity results for scales corresponding to the 10 PD constructs listed in the DSM–5 using data from student, community, clinical, and correctional samples. Initial validation efforts indicated good support for criterion validity with an external PD measure as well as with dimensional personality traits included in the DSM–5 alternative model for PDs. Construct validity results using psychosocial history and therapists' ratings in a large clinical sample were generally supportive as well. Overall, these brief scales provide clinicians using MMPI–2–RF data with estimates of DSM–5 PD constructs that can support cross-model connections between categorical and dimensional assessment approaches.

Acknowledgments

We thank Yossef Ben-Porath and John Graham for granting access to the clinical and correctional data sets used in this study. We are also grateful to the University of Minnesota Press for giving us permission to use the MMPI–2–RF normative sample and for providing research support for the original data collections. This article represents the opinion of Mark H. Waugh and not of the U.S. DOE, ORNL, or UT-Battelle, LLC, on work authored by UT-Battelle, LLC under Contract No. DE-AC05-00OR22725 with the U.S. Department of Energy.

Declaration of interest

Martin Sellbom is a paid consultant to the University of Minnesota Press, publisher of the MMPI–2–RF. He also teaches workshops and webinars on the MMPI–2–RF for remuneration. Finally, although funds were not directly requested for the current project, Dr. Sellbom does receive continuous funding for MMPI–2–RF research from the publisher.

Notes

¹ We elected to start with MWB and SBP scales because they represented the most dominant PD scale sets for the MMPI and MMPI–2, respectively, in the literature and were associated with the largest item pools.

² For copyright and test security reasons, we cannot print the item statements.

³ It also bears mentioning that, despite attempts to control for inconsistent and deviant responding by excluding invalid profiles, the university students scored substantially higher than the normative sample on the Fp-r scale (57T vs. 50T) and in the same range as the outpatient sample with more severe psychopathology (57T vs. 56T). As such, it is quite possible that some of these surprising findings were due to greater levels of deviant responding than actual levels of personality psychopathology.

⁴ About two thirds of clients in this sample had PDF data given that not all clients had three therapy sessions. Because of missing ratings (mostly owing to clinicians marking unknown), the sample sizes ranged from 317 (work problems) to 604 (introverted). Seventeen of the 19 rating variables used in this study had ratings for more than 500 participants.