Construct interpretation of Minnesota Multiphasic Personality Inventory (MMPI) clinical scales: An orthogonal transformation

The implications for personality test construction of the revolution in testing caused by construct validity considerations are outlined, with particular relevance to the assessment of psychopathology. These include (a) substantive definition of constructs; (b) concern for internal consistency reliability as well as generalizability; (c) evaluation of structural relationships among items and scales; (d) suppression of response biases; (e) emphasis on minimum redundancy among scales; (f) evaluation of convergent and discriminant validity of scales and profiles; and (g) evaluation of criterion validity for configurations of scales and profiles, as well as single scales. Benefits are seen as accruing to an increased understanding of psychopathology and higher levels of validity. Prior, and subsequent, to the forthcoming revision of the Minnesota Multiphasic Personality Inventory (MMPI), one approach to realizing some of the aims of construct measurement with an empirically based test is through an orthogonal transformation of the scales. Preliminary results for the extant MMPI clinical scales are reported, yielding evidence of (a) scale independence while retaining high correlations with uncorrected scales, (b) an appropriate pattern of correlations with a separate set of new scales of psychopathology, (c) a possible basis for new item analyses, and (d) freedom from correlations with a putative measure of response bias. Implications of the orthogonal transformation for profile interpretation are discussed.Portions of this paper were presnted at an invited address, 18th Annual Symposium on Recent Developments in the Use of the MMPI, Minneapolis, April 9, 1983. This paper was written while Douglas N. Jackson was distinguished visiting professor at the College of Education, The University of Iowa. This research has been supported by Research Grant 895-84/86 from the Ontario Mental Health Foundation, Research Grant 411-83-0014 from the Social Sciences and Humanities Research Council of Canada, and the Alberta Hospital Edmonton.     

Factor invariance of the Millon Clinical Multiaxial Inventory

Several factor analyses of the Millon Clinical Multiaxial Inventory (MCMI) have resulted in very similar solutions. Interpretation of this consistency is hampered by the fact that the 20 scales of the inventory share items. Overlapping items cause the scales to be linearly dependent and may create structure in the interscale correlation matrix which is separate from the subject response patterns. A factor analysis was performed on the matrix of item-overlap coefficients which describes the underlying artifactual structure of the instrument. Data from two new subject samples were factor analyzed and compared to previously published studies. Similarity coefficients among factors across studies were calculated.     

Millon Clinical Multiaxial Inventory (MCMI) scores with a collegiate sample: Long-term stability and self-other agreement

This study examined (a) the stability of the personality and symptom scales of the Millon Clinical Multiaxial Inventory (MCMI) for a sample of university undergraduates and (b) the correspondence between MCMI scores from self-report versus scores obtained from knowledgeable others who responded by describing the subject rather than themselves. The MCMI was administered to subjects during their freshman year and then again during their senior year. As in clinical populations, stability estimates in this collegiate sample were greater for the basic personality scales than for the symptom scales. Comparison of the results with those from previous studies further showed that the stability coefficients for the collegiate sample were somewhat lower overall than those from treatment follow-up studies with clinical populations. The results also showed that reasonable agreement between self- and others' ratings was obtained on most MCMI scales. The greatest agreement was found for scales that reflect more observable behaviors and relate to an introversion-extroversion dimension, whereas the least agreement was found for scales related to a psychoticism dimension. Lower levels of agreement were also found for scales having a lower mean proportion of items endorsed.This study was sponsored in part by PHS Grant R01 MH31750-01-6, by PHS Grant 5R01 AA06754-01-03, and by funds from the University of Connecticut Research Foundation and Computer Center.     

Stability of the Millon Clinical Multiaxial Inventory among psychiatric inpatients

This study investigated the stability of the Millon Clinical Multiaxial Inventory (MCMI) in a sample of psychiatric inpatients over a longer time period and from different perspectives than previous reliability studies. Test-retest reliability was assessed from three perspectives: correlation/regression, equivalence of dimensional structures across testings upon different admissions, and stability of personality style across testings. Some stability from a correlational perspective was found, with higher stability estimates indicated for the basic personality scales in comparison with the clinical symptom scales which is consistent with Millon's theory. The internal structure of the MCMI was essentially identical across testings. Although source of reinforcement and instrumental or coping style were stable across testings beyond chance, the MCMI profiles of a considerable number of subjects were inconsistent across testings. Correspondingly, clinicians should be cautious in using the MCMI to make categorical decisions regarding personality style with inpatients who are tested at admission and who have had repeated hospitalizations.     

Psychological and sexual functioning among outpatient sexual offenders against children: A Minnesota Multiphasic Personality Inventory (MMPI) cluster analytic study

Multivariate clustering procedures were used to identify homogeneous subgroups of outpatient sex offenders against children (n=110)on the basis of the Minnesota Multiphasic Personality Inventory (MMPI). Results indicated a five-subgroup solution. Two within-normal-limits and two clinically elevated profile subgroups replicated subtypes found in previous cluster analytic studies of child sex offenders. The fifth subgroup appeared similar in MMPI profile pattern and elevation to a previously identified subtype of rapists. Discriminant analyses showed that subgroups differed along dimensions of cognitive disturbance and sexual functioning. Furthermore, levels of psychopathology corresponded with levels of sexual pathology. Results are discussed with reference to previous MMPI cluster analytic studies of sex offenders and theories of sexual aggression.Portions of this paper were presented at the meeting of the Midwestern Psychological Association, 1992, Chicago, Illinois.     

Minnesota Multiphasic Personality Inventory (MMPI)-derived psychopathology subtypes among former prisoners of war (POWs): Replication and extension

Psychopathology and symptom patterns identified among former prisoners of war (POWs) by Sutker, Winstead, Goist, Malow, and Allain (1986) were replicated in an independent sample of 51 former POWs with similar personal backgrounds and military experiences. Data collection instruments included the Minnesota Multiphasic Personality Inventory (MMPI), self-report measures of anxiety and depression, and a structured clinical interview including a POW Trauma Index. Two prototypic MMPI profile patterns were identified using modal profile analysis (Skinner & Lei, 1980). Both were highly similar in shape and elevation to those reported in the previous investigation. Multiprofile-multisample analysis produced prototypic profile patterns which were accurate representations of profiles identified in separate analyses of the derivation and replication samples (r's .96). Representing unique constellations of clinical features, profile subtypes were associated differentially with confinement stress severity, postservice adjustment, and nature and extent of stress-induced symptomatology.     

A response Consistency statistic for the millon clinical multiaxial inventory (MCMI)

Response style in objective psychological testing is an important issue in the reliability and validity of tests as well as in the interpretation of test results. The MCMI provides two response-style indices, the validity scale and the weight factor. The present work presents an additional statistic to assess random response in subjects. The Consistency Coefficient is the correlation between the subjects' endorsement of even and odd items across the 20 MCMI scales. The distributions of 500 patient and 500 randomly generated profiles were compared. Good separation between these distributions was found. The subject data were extremely negatively skewed, whereas the randomly generated data were normally distributed. Data are presented that display positive and negative predictive values, as well as sensitivity and specificity across ranges of prevalence and cut score. These data facilitate the identification of subjects who respond to the MCMI in a random manner so that their scores can be interpreted accordingly.     

Bias and Millon clinical multiaxial inventory (MCMI-II) diagnosis

In the current study, the degree of bias is calculated for each of the personality disorder and clinical syndrome scales of the MCMI-II. In general, most of the MCMI-II scales are prone to only mild or moderate biases. However, the paranoid personality disorder, somatoform, bipolar: manic, thought disorder, and delusional disorder scales are prone to severe biases. When the MCMI-II is utilized to make diagnostic decisions, bipolar and schizophrenic disorders are apt to be grossly underestimated. The implications of bias in diagnostic prevalence rates are discussed and an equation is offered which provides for adjustments to be made when the percentage of positive MCMI-II test results are used to determine prevalence rates for clinical or research populations. The observation is made that imperfect sensitivity and specificity for the MCMI-II scales will result in inaccurate estimates of personality disorders and clinical syndromes when the MCMI-II is used to survey various populations.     

Relationship between the Millon Behavioral Health Inventory and the minnesota multiphasic personality inventory (MMPI) in low-back pain patients

The Millon Behavioral Health Inventory (MBHI) is being used with increasing frequency for the assessment of chronic pain, although there is a relative lack of evidence as to its utility, and prior studies have not examined low back pain. This investigation compared the MBHI to the MMPI in a sample of low-back pain patients and analyzed subgroups of pain patients based upon their MBHI responses. Subjects were 60 patients who had been admitted to outpatient multidisciplinary pain clinics of two Chicago-area hospitals. Patients completed both the MMPI and the MBHI and provided demographic information. Results of correlational analyses indicated strong relationships between the MBHI psychogenic attitude, psychosomatic correlate, and prognostic index scales and the validity scales of the MMPI. The MBHI Pain Treatment Responsivity scale (PP) correlated with 16 of the other 19 MBHI scales. PP did not demonstrate specificity with low back pain patients. The results of both the scale comparisons and the exploratory two-group cluster subgroup analysis support the notion that responses to the MBHI are largely affected by the respondent's tendency to deny psychopathology or to admit emotional distress.     

A General Factor of Personality in the Millon Clinical Multiaxial Inventory-III, the Dimensional Assessment of Personality Pathology, and the Personality Assessment Inventory

A General Factor of Personality (GFP) occupies the apex of the hierarchy in three prominent personality disorder inventories. On the Millon Clinical Multiaxial Inventory-III, a GFP accounted for 41% of the variance in two second-order factors, 31% of the variance in five first-order factors, and 26% of the variance in all 24 scales. On the Dimensional Assessment of Personality Pathology, a GFP accounted for 61% of the variance in six first-order factors and 36% of the variance in all 18 scales. In a cross-validation study of the Personality Assessment Inventory, a GFP accounted for 65% of the variance in two second-order factors, 47% of the variance in five first-order factors, and 27% of the variance in all 18 scales.     

Cross-validation of a minnesota multiphasic personality inventory (MMPI) subscale for the assessment of combat-related post-traumatic stress disorder

We have examined the Minnesota Multiphasic Personality Inventory (MMPI) scores of 80 carefully diagnosed veterans with post-traumatic stress disorder (PTSD) in an attempt to cross-validate the PTSD subscale and three-point MMPI code of Keane, Malloy, and Fairbank (1984). In general, although we replicated Keane and co-workers'cutoff scores, our hit rates were slightly lower than those reported originally.This research was supported in part by NIMH Grant MH-37839 to Dr. Kolb.     

Resolutions to interpretive dilemmas created by the minnesota multiphasic personality inventory 2 (MMPI-2): A reply to Strassberg

We discuss the interpretive dilemmas identified by Strassberg (1991) and propose to resolve these dilemmas by pointing out that some change has indeed been introduced in the MMPI-2 and that change is a necessary requisite for improvement. Changes in administration procedures used with the MMPI-2 normative sample should only improve the quality of the normative data; changes in the way T-scores were developed lead to minimal change in the profile but enable more accurate interpretation of differences between scales on an individual's profile; normative changes have a differential effect on the clinical scales and this is to be expected; and the problem of codetype congruence has been overestimated and overstated. We conclude that the change introduced in the MMPI-2 can only serve to improve the test.     

Failed back surgeries and minnesota multiphasic personality inventory (MMPI) profiles

MMPI profiles were evaluated for 105 prospective surgical patients who had previously undergone surgery or other procedures for treatment of back pain. Patients were classified into groups having undergone zero, one, two, three, or four or more previous surgeries. While all groups demonstrated a characteristicsomatogenic profile, none of the MMPI validity or clinical scales significantly differentiated the groups and there was no relationship between increased number of surgeries and MMPI scale characteristics. These results support the nonoptimistic prognostication of thesomatogenic MMPI profile for surgical intervention for back pain but show no clear relationship of MMPI profile characteristics to degree of experience of previously failed surgery.     

Interpretive dilemmas created by the minnesota multiphasic personality inventory 2 (MMPI-2)

While the new MMPI-2 has the potential to provide more and better clinically relevant information than its predecessor, its introduction also creates several important problems for clinicians and researchers. In particular, the new norms and the use of uniform (rather than linear) T-scores result in the same raw scores on the two tests yielding clinical profiles that can be quite different, both qualitatively and quantitatively. These differences can result in difficulty in MMPI-2 profile interpretation when the user is relying on MMPI interpretive strategies. This paper addresses these concerns and offers some temporary remedial strategies.     

Personality disorder correspondence between the million behavioral health inventory (MBHI) and the minnesota multiphasic personality inventory (MMPI)

The MBHI and MMPI personality disorder scales were analyzed for convergent and discriminant validity. Correlational data demonstrated that six of the eight scales were significantly related, while the remaining two scales approached significance. Further analyses of these data, however, demonstrated that none of the scales correlated significantly better with its convergent scale compared to nonconvergent scales. The MBHI classified significantly more of the sample as personality disordered (93%) compared to the MMPI personality disorder scales (17%). Furthermore, the MBHI tended to describe the sample as falling within the Anxious cluster of personality disorders, whereas the MMPI described them within the Dramatic cluster. Single scale codetype correspondence was found to be 15%, while two-point concordance was 12.5%, indicating very low congruence between personality style codetypes. These two measures do not appear to be measuring the same personality style constructs.     

Diagnostic accuracy of the Millon test computer reports for bipolar affective disorder

The diagnostic accuracy of the Millon Clinical Multiaxial Inventory (MCMI) computer reports was assessed for the sample of 48 patients reliably identified as having bipolar affective disorder. Only 13 of the 48 reports were accurate in classifying the patients as having bipolar affective disorder, and this was significantly (p < .01) less than 50% of the cases. The diagnostic accuracy clearly left much to be desired. Additional research is needed to assess the validity of MCMI diagnostic suggestions for other patient groups.     

Construct validity of the Sivik Psychosomaticism test and test of Operational style: Correlations with four Minnesota Multiphasic Personality Inventory (MMPI) subscales

To evaluate the construct validity (convergent and divergent) of the Sivik Psycho Somaticism test (SPS) and test of Operationality (OPER), Pearson correlation coefficients between SPS scales and subscales, OPER and Minnesota Multiphasic Personality Inventory (MMPI) subscales Hypochondria (Hs), Depression (D), Hysteria (Hy) and Alexithymia (Al) were calculated. Eighty-eight healthy individuals and 285 psychosomatic patients completed the SPS and OPER tests and MMPI; Hs, D, Hy and Al. The results show that most of the SPS subscales and OPER are significantly correlated to several MMPI subscales in both a normal and a psychosomatic population. The results are in concordance with the theoretical hypotheses and confirm the validity of the SPS and OPER constructs.     

Behavioral,psychological, and historical correlates of MacAndrew scale scores among male alcoholics

Male alcoholic veterans (N=194) were divided into four groups on the basis of their scores on the MacAndrew scale (MAC) of the Minnesota Multiphasic Personality Inventory (MMPI) and compared on psychological and sociodemographic measures and self-reported alcohol, drug, and legal histories. Nineteen subjects scored in the nonalcoholic (NA) range of the MAC (i.e., false negatives, MAC23). Remaining subjects, scoring in the alcoholic range of the MAC, were divided into thirds by MAC scores (lowest third, L-MAC=24–28,n=53; middle third, M-MAC=29–31,n=63; highest third, H-MAC>31,n=59). Groups did not differ on age, any important sociodemographic variables, or scores on the Profile of Mood States, State-Trait Anxiety Inventory, Symptom Checklist-90, or Rod and Frame Test. Significant group differences were obtained on several MMPI scales. Lower MAC scores were associated with higher scores onL, K, D, andR, and lower scores onMa. There was also a tendency for subjects with lower MAC scores to score higher onSi. NA subjects began drinking and heavy drinking later than other subjects, although groups did not differ on duration, quantity, or frequency of drinking. Subjects with higher MAC scores more often reported drinking in bars, drinking liquor straight, alcohol-related job disruptions, and previous use of marijuana, hallucinogens, and barbiturates. Subjects with higher MAC scores had more often been arrested, convicted, fined, and jailed for alcohol-related offenses but not for other offenses. Results are examined with respect to MacAndrew's distinction between primary and secondary alcoholics and with respect to the possible relations of MAC to important person variables such as sensation seeking.The authors wish to thank Craig MacAndrew for his helpful critical reading of an early draft of this paper. We also thank Theresa Hilliard, Shirley White, Vicki Brumbelow, and Pam Thompson for editorial assistance and Audrey Jefferson for help with data preparation.     

MMPI Profiles of Liver and Heart Transplant Candidates With and Without Significant Histories of Heavy Alcohol Use

This study compared the Minnesota Multiphasic Personality Inventory (MMPI) profiles of liver and heart transplant candidates with (n = 104) and without (n = 253) significant histories of heavy alcohol consumption. The alcohol and nonalcohol groups had similar overall mean profiles with significant or marginally significant (i.e., T 69) clinical elevations on Scales 1, 2, and 3. However, significantly higher proportions of those in the liver transplant group than in the heart transplant group scored within the clinical range on Scales F, 4, 6, and 8. In addition, the proportions of respondents obtaining clinically elevated scores were significantly higher for the alcohol than for the nonalcohol group on Scale 4 and the MacAndrew Scale; the proportion was higher for the nonalcohol than for the alcohol group on Scale K. The overall similarity of the mean profiles indicates that candidates for liver or heart transplantation with and without histories of heavy alcohol use generally display similar psychological presentations. The observed differences between the liver and the heart transplant groups may reflect impairments in mental status among liver transplant candidates due to metabolic consequences of liver disease. The differences between the alcohol and the nonalcohol groups may be more a reflection of past behavioral patterns than present psychological status.     

Personality disorders and response to outpatient treatment of chronic pain

As part of a comprehensive interdisciplinary evaluation conducted prior to participation in an outpatient chronic pain treatment program, the psychological status of 101 persons was assessed. The majority of participants was found to have a form of personality disorder, determined by conservative cutoff scores applied to their Millon Clinical Multiaxial Inventory (MCMI) profiles. DSM-III-R Cluster C disorders (i.e., Avoidant, Dependent, Obsessive-Compulsive, and Passive-Aggressive) were overrepresented in this sample. Subsequent analyses revealed that personality disorders were related to higher levels of self-reported distress and pain at both the beginning and the end of outpatient treatment. Differential responses to treatment were observed on self-report measures; however, few relations were found between personality disorder and physical therapist ratings of impairment and improvement. Implications for the assessment of personality disorders in outpatient pain treatment programs are discussed and appropriate intervention strategies are considered.