Mapping the MMPI–2–RF Substantive Scales Onto Internalizing, Externalizing, and Thought Dysfunction Dimensions in a Forensic Inpatient Setting

ABSTRACT

Contemporary models of psychopathology—encompassing internalizing, externalizing, and thought dysfunction factors—have gained significant support. Although research indicates the Minnesota Multiphasic Personality Inventory–2 Restructured Form (MMPI–2–RF; Ben-Porath & Tellegen, 2008/2011) measures these domains of psychopathology, this study addresses extant limitations in MMPI–2–RF diagnostic validity research by examining associations between all MMPI–2–RF substantive scales and broad dichotomous indicators of internalizing, externalizing, and thought dysfunction diagnoses in a sample of 1,110 forensic inpatients. Comparing those with and without internalizing diagnoses, notable effects were observed for Negative Emotionality/Neuroticism–Revised (NEGE-r), Emotional/Internalizing Dysfunction (EID), Dysfunctional Negative Emotions (RC7), Demoralization (RCd), and several other internalizing and somatic/cognitive scales. Comparing those with and without thought dysfunction diagnoses, the largest hypothesized differences occurred for Thought Dysfunction (THD), Aberrant Experiences (RC8), and Psychoticism–Revised (PSYC-r), although unanticipated differences were observed on internalizing and interpersonal scales, likely reflecting the high prevalence of internalizing dysfunction in forensic inpatients not experiencing thought dysfunction. Comparing those with and without externalizing diagnoses, the largest effects were for Substance Abuse (SUB), Antisocial Behavior (RC4), Behavioral/Externalizing Dysfunction (BXD), Juvenile Conduct Problems (JCP), and Disconstraint–Revised (DISC-r). Multivariate models evidenced similar results. Findings support the construct validity of MMPI–2–RF scales as measures of internalizing, thought, and externalizing dysfunction.

Acknowledgments

The statements and opinions in this article are those of the authors and do not constitute the official views or the official policy of DSH-Patton, the California Department of State Hospitals, or the State of California. We would like to express our appreciation to Mr. Gareth Meeson, who assisted in data cleaning efforts, and to Harry Oreol for his support of the research program at Patton State Hospital.

Funding

This research was supported by a grant to David M. Glassmire and Danielle Burchett from the MMPI–2–RF publisher, the University of Minnesota Press–Test Division, which holds the copyright to the MMPI–2–RF. It was also supported by the California State University, Monterey Bay Undergraduate Research Opportunities Center, the U.S. Department of Education Ronald E. McNair Scholars Achievement Program (#P217A120262), and a U.S. Department of Education Hispanic-Serving Institution (HSI) Science, Technology, Engineering, and Mathematics (STEM) and Articulation Program Individual Development Grant (#P031C110121). Anthony M. Tarescavage receives research funds unrelated to this study from the University of Minnesota Press–Test Division.

Notes

¹ Hoelzle and Meyer (2008) noted five factors within the Restructured Clinical scales, measuring depressive (RCd, RC2), psychotic (RC6, RC8), externalizing (RC4, RC9) and somatic (RC1) and cynicism (RC3) domains.

² Or the longer MMPI–2 version that can be used to score the MMPI–2–RF, as discussed further in the Procedure section.

³ Of note, although some research has demonstrated evidence for a fourth somatoform factor (e.g., Kotov, Ruggero, et al., 2011), a construct assessed by the MMPI–2–RF, the limited number of such diagnoses in this setting did not allow for the creation of a somatoform criterion group in this study.

⁴ All substantive scales in each scale set were examined, including those we did not hypothesize would be associated with the criterion, in an effort to examine MMPI–2–RF discriminant validity.

⁵ Of note, odds ratios for each predictor in each model are presented in Table 2. As an example, the odds ratio for EID in the prediction of internalizing diagnosis is 1.041. This indicates that, holding other scales in the model constant, a 1-point T score increase in EID is associated with 4.1% increased odds of having an internalizing diagnosis. For greater interpretive utility, it is possible to examine the increased odds for a 10-point increase (i.e., 1 SD increase in the MMPI–2–RF normative sample) by calculating OR₊₁₀ = e^(B*10) = e^(.040*10) = 1.492. Thus, a 10-point increase in EID T score is associated with 49.2% increased odds of having an internalizing diagnosis. Similarly, the odds ratio associated with a 10-point increase in MEC score is e^(−.022*10) = 0.803, indicating a (1.00 – 0.803) 19.7% decrease in the odds of having an internalizing disorder, holding AES score constant.