Interpreting Supplementary Scales of the MMPI-2
This overview of the MMPI-2 supplementary scales will be organized into four groupings of scales: generalized emotional distress scales (Welsh Anxiety [A], College Maladjustment [Mt], and Post Traumatic Stress Disorder–Keane [PK]); control/inhibition and dyscontrol/dysinhibition scales (Welsh Repression [R], Hostility [Ho], and MacAndrew Alcoholism–Revised [MAC-R]); alcohol/drug scales (MacAndrew Alcoholism–Revised [MAC-R], Addiction Admission [AAS], Addiction Potential [APS], and Common Alcohol Logistic–Revised [CAL-R]); and the Personality Psychopathology Five scales (PSY-5: Harkness, McNulty, & Ben-Porath, 2002). Given the limited amount of space, the less frequently used supplementary scales covering general personality dimensions (Dominance [Do], Over-Controlled Hostility [O-H]; Social Responsibility [Re]) and gender role scales (Gender Role–Feminine [GF] and Gender Role–Masculine [GM]) will not be discussed. Information on all of these supplementary scales can be found in Friedman, Lewak, Nichols, and Webb (2001), Graham (2006), and Greene (2011). Clinicians should keep in mind that a general style for individuals to maximize or minimize their reported symptoms will have a significant impact on the elevation, or lack thereof, for all of the MMPI-2 supplementary scales, as well as the standard validity and clinical scales and content scales. It will be assumed in discussing the supplementary scales that individuals have endorsed the items in an accurate manner.
(p. 125) Generalized Emotional Distress Scales
Factor-analytic studies of the MMPI-2 clinical scales have consistently identified two factors that are variously labeled and interpreted. The first factor is generally acknowledged to be a measure of generalized emotional distress and negative affectivity, and Welsh developed his Anxiety (A) scale to measure this factor. The major content areas in the Welsh Anxiety (A) scale are (1) problems in attention and concentration; (2) negative emotional tone and dysphoria; (3) lack of energy and pessimism; (4) negative self-evaluation and hypersensitivity; and (5) obsessions and ruminations. There are 10 to 20 other scales in the MMPI-2 that measure this factor of generalized emotional distress and negative affectivity, all of which have high positive correlations with the A scale: the clinical scales 7 (Pt: .95) and 8 (Sc: .90); the content scales Work Interference (WRK: .93), Depression (DEP: .92), Anxiety (ANX: .90), Obsessions (OBS: .89), and Low Self-Esteem (LSE: .87); the restructured clinical (RC) scales Demoralization (RCD: .94) and Dysfunctional Negative Emotions (RC7: .90) and the supplementary scales Post Traumatic Stress Disorder–Keane (PK: .93), College Maladjustment (Mt: .93), and Marital Distress Scale (MDS: .79). There also are a number of MMPI-2 scales that have high negative correlations with the A scale and as such are simply inverted measures of generalized distress: Ego Strength (Es: −.83) and K (Correction: −.79). All of these scales can be characterized as generalized measures of emotional distress with little or no specificity despite the name of the scale, and there are little empirical data to support any distinctions among them.
Control/Inhibition and Dyscontrol/Dysinhibition Scales
The second factor identified in these factor-analytic studies of the MMPI-2 clinical scales is a measure of control and inhibition, and Welsh developed his Repression (R) scale to measure this factor. The major content areas in the Welsh Repression (R) scale are the denial and suppression of and/or constriction and inhibition of interests either positive or negative in (1) health and physical symptoms; (2) emotionality, violence, and activity; (3) family and relationship problems; (4) social dominance and social participation; and (5) personal and vocational pursuits. There are 5 to 10 other scales in the MMPI-2 that measure this factor of control and inhibition, but the pattern of correlations with the R scale is much more variable and smaller than found with the A scale: the clinical Scale 9 (Ma: .43), the content scale Antisocial Practices (ASP: .36), the restructured clinical (RC) scale Hypomanic Activation (RC9: .60), the supplementary scales MacAndrew Alcoholism–Revised (MAC-R: .50) and Social Responsibility (Re: .38), and the PSY-5 scales Aggression (AGGR: .49) and Disconstraint (DISC: .43). The specific correlates of the second factor will be a function of the scale that is used to define it, but it is evident that this group of MMPI-2 scales is characterized by significant dyscontrol or dysinhibition associated with acting out or externalization of psychopathology.
Conjoint interpretations of the first two factors of the MMPI-2 (generalized emotional distress and control/inhibition) provide a succinct approach for how individuals are coping with the behaviors and symptoms that led them to treatment (see Greene, 2011, Table 7.5, p. 269). The A scale provides a quick estimate of how much generalized emotional distress the individual is experiencing, and the R scale indicates whether the individual is trying to inhibit or control the expression of this distress. It is particularly noteworthy in a clinical setting when the A scale is not elevated (T < 50) because it signifies that the individual is not experiencing any distress about the behaviors and symptoms that usually leads someone else to refer them to treatment. Similarly low scores (T < 45) on the R scale suggest that the individual has no coping skills or abilities to control or inhibit the overt expression of their distress. When one of these two scales is elevated significantly and the other scale is unusually low, clinicians should give serious consideration to the (p. 126) hypothesis that the individual is maximizing (A > 75; R < 45) or minimizing (A < 45; R > 60) his or her report of psychopathology. It is particularly pathognomonic when both the A and R scales are low (T < 50), a pattern that is seen in chronic, ego syntonic psychopathology.
Alcohol and Drug Scales
The alcohol and drug scales on the MMPI-2 can be easily subdivided into rationally derived, or direct, measures (Addiction Admission [AAS] and Common Alcohol Logistic–Revised [CAL-R]) and empirically derived, or indirect, measures (MacAndrew Alcoholism–Revised [MAC-R] and Addiction Potential [APS]). These four alcohol and drug scales contain 111 different items, 96 of which are found on only one of the four scales. These different methodologies yielded very different item groupings on these four scales that can be seen in the low positive correlations among them: MAC-R with AAS .48, APS .29, and CAL-R .32; and AAS with APS .34; CAL-R .46. Consequently, the manifestations of alcohol and drug abuse will differ in specific individuals depending upon which scale is elevated.
The MAC-R scale is best conceptualized as a general personality dimension. Individuals who produced elevated scores (T scores > 64) on the MAC-R scale are described as being impulsive, risk taking, and sensation seeking, and they frequently have a propensity to abuse alcohol and/or stimulating drugs. They are uninhibited, sociable individuals who appear to use repression and religion in an attempt to control their rebellious, delinquent impulses. They also are described as having a high energy level, having shallow interpersonal relationships, and being generally psychologically maladjusted. Low scorers (T scores < 45) are described as being depressed, inhibited, overcontrolled individuals, who also may abuse substances, but in a different manner. If they abuse substances, they will prefer alcohol or sedating drugs. Once the MAC-R scale is understood as a general personality dimension for risk taking versus risk avoiding, the fact that mean scores vary drastically by codetype makes sense. For example, in men, the mean T score on the MAC-R scale in a 4-9/9-4 (risk-taking) codetype is 63.4 and in a 2-0/0-2 (risk-avoiding) codetype is 42.5 (see Greene, 2011, Appendix A), a difference of over two standard deviations. There are a number of issues that must be kept in mind when interpreting the MAC-R scale: Men score about 2 raw-score points higher than women across most samples, which indicates that different cutting scores are necessary by gender; there is not a single, optimal cutting score with raw scores anywhere from 24 to 29 being used in different studies; cutting scores appear to be influenced by a number of factors, so clinicians need to begin to determine empirically the best cutting score for their specific treatment facility to optimize the percentage of clients correctly classified as substance abusers; clinicians need to be very cautious in using the MAC-R scale in non-White ethnic groups, if it is used at all; classification accuracy decreases when clinicians are trying to discriminate between substance abusers and non-substance-abusing, clinical clients, which is a frequent differential diagnosis; classification accuracy may be unacceptably low in medical samples; and the MAC scale is a general measure of substance abuse that is not specific to alcoholism.
The Addiction Potential Scale (APS) consists of 39 items that differentiated among groups of male and female substance-abuse patients, normal individuals, and clinical clients. Individuals with elevated (T > 64) scores on the APS scale are generally distressed and upset, as well as angry and resentful. They also are concerned about what others think of them, a concern that is not evident in individuals who elevate the MAC-R scale. Low scorers (T < 45) on the APS describe themselves in relatively positive terms. They are not distressed or angry. If they are abusing substances, they either are not experiencing or not reporting any negative consequences. The APS scale appears to be more accurate at discriminating between substance-abuse clients and clinical clients than is the MAC-R scale. The APS scale also tends to be less gender biased than the MAC-R scale and to be less codetype sensitive. For example, in men, the mean T score on APS in a 4-9/9-4 codetype is 55.9 (p. 127) and in a 2-0/0-2 codetype is 48.5 (see Greene, 2011, Appendix A), a difference slightly over one-half of a standard deviation.
The Addiction Admission Scale (AAS) consists of 13 items, 9 of which are directly related to the use of alcohol and drugs. Clinicians should review the clinical history and background of any individual who elevates the AAS scale (T > 64) because of the explicit nature of the items and the fact that four or more of these items have been endorsed in the deviant direction to produce this elevation. The AAS scale typically performs better at identifying individuals who are abusing substances than less direct measures such as the APS and MAC-R scales, even though the items are face valid, allowing individuals not to report the substance abuse if they desire to do so. Weed, Butcher, and Ben-Porath (1995) have provided a thorough review of all MMPI-2 measures of substance abuse.
The Common Alcohol Logistic (CAL) scale because was developed out of concern that existing MMPI alcohol scales lacked adequate positive predictive power given the low base rate or prevalence of alcohol-related problems in general medical settings. Gottesman and Prescott (1989) raised similar concerns about the MAC-R scale in clinical clients. The 33 items for the CAL scale were identified and the item weights were assigned by using logistic regression in large samples of alcoholic clients, medical clients, and normal individuals. Malinchoc, Offord, Colligan, and Morse (1994) revised the CAL scale for the MMPI-2 by dropping the six items on the CAL scale that were not retained on the MMPI-2. They recomputed the item weights using logistic regression on similar groups of patients, and the resulting 27 items became the CAL-R scale that is appropriate for use with either the MMPI or MMPI-2. They did not use the MMPI-2 item pool in this revision, so it remains to be seen whether any of the new MMPI-2 items, particularly the items asking about alcohol and drug abuse, would have been selected for inclusion on the scale. High scorers (T scores of 65 or higher) on the Common Alcohol Logistic–Revised (CAL-R) scale have used alcohol excessively and they feel alienated from others and members of their family. They do not report physical symptoms as a consequence of their use of alcohol. The CAL-R scale appears to be particularly useful to identify substance abuse in medical settings, no doubt reflecting the context in which the scale was developed.
Although the focus of this section is on alcohol and drug scales, it is important to note that there are a number of specific MMPI-2 items related to alcohol and drug use (264, 489, 511, 544) that warrant further inquiry any time they are endorsed in the deviant direction. Most of these items are phrased in the past tense so the clinician cannot assume without inquiry whether the alcohol and drug use is a current or past event.
Personality Psychopathology Five Scales
Harkness and McNulty created a five-factor model called the Personality Psychopathology Five (PSY-5) to aid in the description of normal personality and to complement the diagnosis of personality disorders. Using replicated rational selection, Harkness and McNulty identified five factors within 60 descriptors of normal and abnormal human behavior: Aggressiveness (AGGR), Psychoticism (PSYC), Disconstraint (DISC), Negative Emotionality/Neuroticism (NEGE), and Introversion/Low Positive Emotionality (INTR) (cf. Harkness, McNulty, Ben-Porath, & Graham, 2002). The AGGR scale assesses offensive aggression and possibly the enjoyment of dominating, frightening, and controlling others, and the lack of regard for social rules and conventions. The PSYC scale assesses the cognitive ability of the individual to model the external, objective world in an accurate manner. Persons who are low on the PSYC construct can realize that their model is not working and accommodate or revise the model to fit their environment. Although the PSYC scale has its largest correlations with Scales 8 (Schizophrenia [Sc]: .78) and Bizarre Mentation (BIZ: .90), it appears to be measuring a general distress factor, much like the NEGE scale. The DISC scale assesses a dimension from rule following versus rule (p. 128) breaking and criminality. The DISC scale is not correlated to most of the other MMPI-2 scales and, thus, would appear to have the potential to contribute additional information when interpreting the MMPI-2. The largest correlations of the DISC scale are with Scales 9 (Hypomania [Ma]: .40), MacAndrew Alcoholism–Revised (MAC-R: .51), Addiction Admission (AAS: .52), and Antisocial Practices (ASP: .57). The NEGE scale assesses a broad affective disposition to experience negative emotions focusing on anxiety and nervousness. The NEGE scale is another of the numerous markers for the first factor of general distress and negative emotionality on the MMPI-2. The INTR construct assesses a broad disposition to experience negative affects and to avoid social experiences. Although the INTR scale generally has its largest correlations with MMPI-2 markers for the first factor, the INTR scale is a measure of anhedonia that is suggestive of rather serious psychopathology. Such an interpretation of the INTR scale is particularly likely when the NEGE scale is not elevated significantly.
The PSY-5 scales are another potential source of information for the clinician in interpreting the MMPI-2 profile. Research that demonstrates their usefulness in patients with personality disorder diagnoses is needed. Until such information is available clinicians are cautioned to interpret them very conservatively.
The MMPI-2 supplementary scales should be scored and interpreted routinely as a valuable source of additional information that is not readily available in the standard validity and clinical scales or the content scales. For example, the conjoint interpretation of the A and R scales provides a quick insight into how individuals are experiencing and coping with their psychopathology that brought them to treatment. In addition, the information on alcohol and drug use can only be inferred indirectly from the MMPI-2 clinical and content scales, while this information is available both directly and indirectly in the supplementary scales. The information provided by the supplementary scales is invaluable in the treatment-planning process.
References and Readings
Friedman, A. F., Lewak, R., Nichols, D. S., & Webb, J. T. (2001). Psychological assessment with the MMPI-2. Mahwah, NJ: Erlbaum.Find this resource:
Gottesman, I. I., & Prescott, C. A. (1989). Abuses of the MacAndrew MMPI alcoholism scale: A critical review. Clinical Psychology Review, 9, 223–242.Find this resource:
Graham, J. R. (2006). MMPI-2: Assessing personality and psychopathology (4th ed.). New York: Oxford University Press.Find this resource:
Greene, R. L. (2011). The MMPI-2: An interpretive manual (3rd ed.). Boston, MA: Allyn & Bacon.Find this resource:
Harkness, A. R., McNulty, J. L., Ben-Porath, Y. S., & Graham, J. R. (2002). MMPI-2 Personality Psychopathology Five (PSY-5) scales: Gaining an overview for case conceptualization and treatment planning. Minneapolis: University of Minnesota Press.Find this resource:
Malinchoc, M., Offord, K. P., Colligan, R. C, & Morse, R. M. (1994). The common alcohol logistic–revised scale (CAL-R): A revised alcoholism scale for the MMPI and MMPI-2. Journal of Clinical Psychology, 50, 436–445.Find this resource:
Weed, N. C., Butcher, J. N., & Ben-Porath, Y. S. (1995). MMPI-2 measures of substance abuse. In J. N. Butcher & C. D. Spielberger (Eds.), Advances in personality assessment (Vol. 10, pp. 121–145). Hillsdale, NJ: Erlbaum.Find this resource: