Chapter Seven

CHAPTER 7
Interpreting the MMPI-2 Supplementary Measures

Because the standard scales of the MMPI-2 focus primarily on psychopathology, they do not necessarily address all of the personality characteristics/problems that are of special interest to clinical psychologists. This is where the supplemental scales come in – they are much narrower and more specific in scope than the standard scales. Because of this, they can be used to augment the standard MMPI-2 measures.

MacAndrew Alcoholism Scale
Originally developed to assess alcohol abuse in clinical settings.
Differentiates people who have alcohol abuse problems from people who have psychiatric problems not relating to alcohol.
Modified from 51 items to 49, because two items were too obvious, and it was assumed that many alcoholics would not admit to abusing alcohol. Therefore, these items would distort the results.
Raw score cutoff of 26 for men and 24 for women (up from 24 and 22, respectively). If these lower cutoffs are used, the scale is less effective.
Empirical research supports the use of this scale as an assessment device for addiction problems
T = 60 is considered suggestive of a substance abuse problem.
More a measure of addiction proneness rather than specifically an alcohol use/abuse scale.
Items selected based on how they differentiated between alcoholics and non-alcoholics
New scale constructed by deleting objectionable items and replacing them with ones from the new MMPI-2 item pool
New scale has the same power as the old version in differentiating between alcoholics and non-alcoholics.
Interestingly men who had higher arrest records had MAC-R scores that were nearly identical to men who were problem drinkers – a finding possibly related to the fact that these men’s legal problems were related to an underlying substance-abuse problem.
High scores on the MAC-R related to a general recklessness, regardless of gender.
Recent studies have found that the MAC-R is an effective screening device for detecting potential substance abuse problems.
This scale assesses behaviors that are relevant to determining alcohol or drug problems.

*** See p. 157-159 for case study – it’s pretty interesting, but can’t really be condensed from it’s original form, so it’s pointless to try. Useful because it provides an illustration of the scale***



Addiction Potential Scale
Originally designed as a measure of the personality factors underlying the development of addictive disorders.
Empirically derived by selecting items that differentiated alcoholics and drug abuser from psychiatric patients and normals.
Contains 39 items, 9 of which overlap with the MAC
This scale outperformed the MAC-R and another MMPI addiction scale (Substance Abuse Proneness Scale) in discriminating groups of substance abusers from psychiatric patients and normals.
It has been found that APS significantly discriminates substance-abusing psychotherapy patients from those who do not abuse drugs or alcohol.
See Table 7-2, p. 160 for interpretative guidelines

Addiction Acknowledgment Scale
Developed as a measure of willingness to acknowledge problems with alcohol or drugs.
Provides a psychometric comparison of acknowledged alcohol or drug problems
Contains 13 items derived by using a combined rational-statistical scale construction strategy.
Initially items that obviously addressed substance abuse problems were selected from the MMPI-2 pool, then this scale was correlated with the other MMPI-2 items to see if any other items in the pool were significantly associated with it.
The scale was “purified” by examining the alpha coefficients, keeping items that improved scale homogeneity.
The AAS has been cross-validated in several recent studies.
The AAS has been found to be the most powerful discriminator (more than the MAC-R and APS) of substance abusing clients in an outpatient psychotherapy study.
AAS adds significant incremental validity over the MAC-R in assessing substance abusing clients in an outpatient sample
See Table 7-3, p. 160 for interpretative guidelines

*** See p. 161-162 for case study – it’s pretty interesting, but can’t really be condensed from it’s original form, so it’s pointless to try. Useful because it provides an illustration of the scale***









Marital Distress Scale
Marital therapists tried to understand marital maladjustment by looking at the personality profiles of husbands and wives.
Prior studies on couples in marital distress have focused on clinical scale and profile differences between distressed and “normal” couples.
The Pd scale has been found to be associated with marital disturbance.
The MMPI-2 FAM content scale is also significantly related to marital distress.
Neither of the above scales were developed for or specifically related to marital distress – they are associated with family problems in general.
The MDS (Marital Distress Scale) is a 14-item scale developed specifically for the assessment of marital distress.
The MDS is more highly related to measured marital distress than the Pd or the FAM scale.
See Table 7-4, p. 163 for interpretative guidelines

*** See p. 163-166 for case study – it’s pretty interesting, but can’t really be condensed from it’s original form, so it’s pointless to try. Useful because it provides an illustration of the scale***

Hostility Scale
Developed to identify people who could work harmoniously within a group, could establish rapport with others, and who could maintain group morale.
Developed the scale by comparing samples of school teachers that were formed on the basis of their ability to get along with others.
People who score high (T > 60) on the Ho scale (Hostility scale) appeared to have difficulty with anger and judged other people in a negative light, and tend to be viewed as overtly hostile in interpersonal situations and generally negative in their approach to others.
Associated with the development of atherosclerosis. Men who developed coronary heart disease were found to score high on this scale.
Used in a lot of health psychology research to examine the relationship between hostility and health behavior. Specifically, the scale has been shown to assess the personality factors of Type A personalities that appear to be associated with the development of coronary heart disease.
MMPI-2 Ho scale items are the same as those on the original MMPI, except that 9 items were edited slightly to make them more readable.
Useful to clinicians in appraising personality factors central to a hostile, competitive lifestyle.
See Table 7-5, p. 167 for interpretative guidelines

*** See p. 163-166 for case study – it’s pretty interesting, but can’t really be condensed from it’s original form, so it’s pointless to try. Useful because it provides an illustration of the scale***

OTHER SCALES

Dominance Scale
Developed to assess personality characteristics of social dominance.
High and low dominant individuals were defined by asking their peers to identify the subjects as either passive or dominant in social relationships.
Measures personality attributes such as comfort in social relationships, self-confidence, possessing strong opinions, etc.
Individuals who score high on this scale are generally viewed as dominant in social situations.
Use in personnel screening to determine if a given individual has any personality characteristics that may interfere with job performance.
See Table 7-6, p. 170 for interpretative guidelines

Social Responsibility Scale
Developed as an assessment of an individual’s sense of responsibility toward others.
People who are high and low “responsible” as identified by peer or teacher ratings served as criterion groups.
Originally composed of 32 items – two items were deleted upon revision, leaving a total of 30 items.
High scores (T> 60) suggest that an individual is dependable and willing to accept the consequences of his/her behavior.
Low scores (T > 40 – not sure if this is a typo and if they really meant “<”, so probably should clarify) suggest that an individual is unwilling to assume responsibility and may not have accepted societal values or conduct.
Widely used in personnel screening as an index of positive personality characteristics.
See Table 7-7, p. 171 for interpretative guidelines

Posttraumatic Stress Disorder Scale
Developed to assess the presence of PTSD symptoms in military veterans.
Compared veterans diagnosed with PTSD to ones who had other psychiatric problems.
Contains 49 items with a broad range of somatic and psychological symptoms.
Appears to be highly related to other MMPI indices of anxiety, such as the Pt and A scales.
Combat veterans and others who have experiences a catastrophic event who score high on this scale are likely to have PTSD symptoms, including anxiety, depression, emotional turmoil, sleep disturbance, intrusive thoughts, etc.
See Table 7-8, p. 171 for interpretative guidelines






Ego Strength Scale
Developed as a potential measure of the ability to benefit from a psychotherapeutic experience
Developed by contrasting groups of successful and unsuccessful patients in psychotherapy.
Research suggests that it actually measures “ability to withstand stress” more that potential for therapeutic success.
Originally contained 68 items – 12 were deleted during MMPI revision
The extent to which the MMPI-2 version of this scale assesses the characteristics measured by the original scale has not been determined
So…..it should be considered experimental until empirical research has further documented its predictive power.
See Table 7-9, p. 172 for interpretative guidelines


Anxiety Scale
Developed as a measure of the first and largest factor dimension in the original MMPI, which also defines the first factor of the MMPI-2
39 items, one of which is endorsed as False.
Contains items that assess general maladjustment or emotional upset.
Not widely used as a clinical assessment measure, but is valuable in research as a marker variable of the first MMPI factor.
The info provided by this scale is available through other MMPI-2 measures, especially the Pt scale, which is highly correlated with A (.75).
Researchers can evaluate proposed new measures by determining their relationship to the A factor.
Individuals who score high on the A scale are endorsing symptoms of anxiety, tension, inability to function, and lack of efficiency in managing everyday affairs and admitting numerous psychological symptoms
See Table 7-10, p. 173 for interpretative guidelines


Repression Scale
Like the A scale the R scale was derived through factor analysis.
Marks the second major factor in the MMPI item pool = overcontrol or denial of conflict.
Composed of 37 items that assess the tendency to deny problems.
Also measures the tendency for people to respond False to everything, even to somewhat neutral items.
Scale content addresses overcontrol and defensive reliance on denial and repression.
People who score high on this scale tend to be uninsightful, overcontrolled, and socially inhibited, considered emotionally constricted, bland, and nonspontaneous.
High scorers tend to deal with conflict by avoidance rather than direct action. They report few psychological problems, tending rather to view themselves as problem-free compared to other people.
See themselves as conventional and reserved in relationships.
High scorers are viewed by others as overly conservative in actions and behavior.
See Table 7-11, p. 173 for interpretative guidelines.


The Personality Psychopathology 5 (PSY-5) Scales
Personality can be understood by FIVE characteristics or traits.
Referred to as the Five Factor Model or the “Big Five”
The people to established this view looked though dictionaries to obtain a multitude of adjective by which people can be described
This multitude was reduced, via factor analysis, to 5 main factors: Agreeableness, extraversion, conscientiousness, emotional stability vs. neuroticism, and openness to experience.
The items developed to mark the 5 factors were composed of “normal” range personality characteristics rather than psychopathology.
Later….. a model that had content validity for assessing psychopathology was developed.
They developed their descriptors by looking at the DSM-III rather than through common adjectives.
Then, they used the MMPI-2 item pool to select items that matched the 5-Factor psychopathological content model, and found a sufficient number of items to serve as test stimuli for the PSY-5 model.
See Table 7-12, p. 174 for item-scale membership of the PSY-5 scales.
The PSY-5 scales are as follows:
1) Aggressiveness – assesses the potential for offensive and instrumental aggression.
2) Psychoticism – measures reality contact or distorted views of the social and object world.
3) Disconstraint – adapted from the “constraint” construct, and measured elements of risk aversiveness, desire for plans and order rather than impulsive action, and traditional morality. Items assess rule-following vs. rule-breaking behavior and criminal behavior.
4) Negative Emotionality/Neuroticism – a broad affective disposition of unpleasant emotions, particularly anxiety, nervousness, and guilt, leading to internal suffering.
5) Introversion (Lack of Positive Emotionality) – broad affective disposition related to the difficulty of experiencing positive affect, the desire to avoid social experiences, and the lack of energy to pursue goals and be engage in life’s tasks.

PSY-5 scales were shown to have cross-measure consistency with the NEO-PI scales.
Aggressiveness appears to combine some aspects of low Agreeableness and high Extraversion
Constraint may be characterized by high Agreeableness and high Conscientiousness
Psychoticisn was positively related to openness.
It appears that the PSY-5 and the NEO-PI assess overlapping but not identical sets of constructs.
Right now, the PSY-5 scales have not been used much in clinical interpretation of the MMPI-2, but are promising new measures warranting further study.


Useful Indexes for the MMPI-2
A number of indexes that combine MMPI scales have been developed for special interpretative purposes
These are either clinically or actuarially derived measures and serve to give the interpreter a valuable perspective on scale relationships.
The following two are the more useful indexes…..

Megargee Classification System for Criminal Offenders
10 types of criminal offenders were identified in the Florida State Correctional System by using their MMPI test patterns and related a number of demographic variables and prison behaviors to these profile groupings.
Widely used
Empirical support for the system
The 10 Megargee types appear to replicate across different inmate samples (e.g. maximum security, medium security, etc). No well-defined demographic characteristics are associated with the different types. The only exceptions are the CHARLIE and HOW types.
CHARLIE = class of inmate most consistently associated with psychological maladjustment and poor adjustment to prison.
HOW = fairly consistently related to adjustment problems.
Correlates of the types, though, have not been consistently found to generalize across settings.
Research so far does not support the stability of the types over time….however whether changes reflect unreliable typology or actual changes in inmate personality, coping style, behavior, and so on remains an open question
Caution should be exercised in making administrative or clinical decisions based on the Magargee typology, but the system could provide useful hypotheses when supplemented with other information.
The Magargee classification system has been revised to accommodate MMPI-2 scale scores.
The Chronic Pain Typology
The MMPI has been most widely used personality measure for investigating the personality characteristics and symptomatic behaviors of chronic pain patients.
It was demonstrated that four main clusters of MMPI types account for a significant percentage of pain patients.
Pain patients from a wide variety of treatment settings appeared to fall into the four groups, which are referred to as P, A, I, N subtypes.
Research has shown that these profile types represent distinctly different problem and personality types in terms of symptoms, treatment amenability, and outcome.
The PAIN classification guidelines are overly conservative in assigning patients to a cluster type.

Critical Items
This approach involves using individual MMPI items as signs or pathognomic indicators of pathology, specific content themes, or special problems the patient is experiencing.
Developed initially by selecting items believed to reflect particular problems of interest to the clinician.
First empirically valid set of critical items was designed to differentiate between patients who were experiencing a crisis requiring hospitalization from other patients not experiencing that particular crisis
Items that differentiates these two groups were found to have both empirical and content validity, and to have considerable clinical utility when used as clues to special problems.
Two other critical items sets are the alcohol-drug crisis set and the depressed-suicidal set.

How are Critical Items Used in Clinical Assessment?
Used clinically and impressionistically to suggest possible interpretative hypotheses about the individual’s current problems, beliefs, or attitudes.
For example, if you are seeing a patient who is depressed, you may want to specifically look at items that ask the individual to report their thoughts on suicide…..ex.
506. I have recently considered killing myself (T)
520. Lately I have thought a lot about killing myself (T)
524. No one knows it but I’ve tried to kill myself (T)
Shouldn’t use these as psychometric indicators or scales as they are much less reliable than are groups of items or scales.
Of greatest value as a suggested hypothesis to follow up in clinical interview or in further scrutiny of the assessment data.

Cautions about other Supplementary Measures
“Blind Empiricism” promoted the development of hundreds of scales, many based on samples of convenience that has little underlying conceptual rationale.
Many of these scales were not sufficiently cross-validated before being used.
Some of these experimental MMPI scales, developed to study particular problems and meant for somewhat limited application, have been expanded and used for purpose very different from those their originators intended.
Should not discourage the development of new empirical scales but to caution users that the development and naming of a scale does not necessarily imply that it satisfactorily assesses the behaviors it purports to measure.
Many MMPI scales have been developed that do not consistently perform as proposed and may yield misleading results.

The Subtle Keys
The most widespread misconception about the MMPI is that several clinical scales contain “subtle” items that allow clinicians to assess individuals without their being consciously aware that they are providing important personal information in the assessment.
Several major points to consider when evaluating the utility of subtle items as predictors of psychopathology and as validity indicators.
1) lack of content relevance in subtle items is that they are imperfectly related to the criterion and were actually selected by chance as a result of incomplete validation procedures. Cross validation should have removed chance items, leaving only those items that are valid indicators of the characteristic being assessed.
2) In evaluating the subtle subscales, they have not been shown to be valid measures of the types of psychopathology they purportedly measure. Subtle items do not predict behaviors as clearly as the obvious items on the scales do.
3) Subtle items give the interpreter a way to evaluate invalid or unusual response attitudes
4) The subtle items are not able to correct for intentional deception on the part of the test taker
The MMPI-2 obvious-subtle keys have been discontinued by the test publisher and are not available either in computer-scoring programs or for hand scoring.

Developing and Evaluating New Scales for the MMPI-2 and MMPI-A
Likely that new scales will be develop for the MMPI-2 and the MMPI-A.
This would be desirable since both the MMPI-2 and the MMPI-A contain some new items that would allow for an expanded assessment and could improve the existing clinical scales.
A scale should be used because it performs a particular assessment better or more efficiently than other scales or provides valuable info not accessible through another assessment approach.
Guidelines for the evaluation of potential MMPI-2 and MMPI-A scales:
1) The construct under study is well-defined
2) The item pool is relevant for the construct being assessed
3) The research design includes cross-validation
4) The scale has appropriate statistical properties, including:
a) adequate internal consistency if the scale is desirable as a measure of a single dimension.
b) A meaningful factor structure, which should be reported
c) If the proposed measure is an empirical scale, it is not required to have split-half reliability. But it should possess other relevant psychometric properties, such as test-retest reliability.
5) All scales, even those developed by rational or internal consistency methods, should have demonstrable validity
6) Uses are explored and demonstrated
7) Correlations between the proposed scale and other well-established MMPI-2 and MMPI-A measures are presented.

***there appears to be a case study (only half a page long, p. 184 and 185) that continues from Chapter 4 and 6. Have a look if you are interested.