Graham, J. R., MMPI-2: Assessing personality and psychopathology, 5th edition. H7: Restructured Clinical (RC) and Personality Psychopathology Five (PSY-5) Scales. Pagina 156-181

Development of the RC scales

The first step in constructing the RC scales was to develop a general scale labelled Demoralisation. Removal of this general factor from the clinical scales would result in a set of restructured scales that would be less intercorrelated and have greater discriminant validity.

Demoralisation is equivalent to the pleasant-unpleasant dimension of self-reported affect.

The next step was designed to identify the core component of each original scale with the general demoralisation factor removed. Factors were extracted and rotated in a manner that yielded a clear demoralization factor, which included all of the provisional Demoralisation scale items as well as items in the clinical scales that are primarily demoralisation markers, and a second factor that could be considered the substantive core for a clinical scale.

The third step involved identifying items with high loadings on the factor representing the core of each to serve as a ‘seed’ scale for each RC scale. A given seed scale was assigned those items that had the highest loadings on the scale’s core factor and that did not have salient loadings on the demoralisation factor. Overlapping items were removed, as were items that detracted from the internal consistency of a seed scale.

The 12 seed scales (demoralisation plus one for each original clinical scale except scale 5) were correlated with all other MMPI-2 items in four clinical samples. Items were selected for the RC scales if they had high correlations with a particular seed scale and low correlation with other seed scales. Items were deleted from a scale if they did not contribute to the scale’s internal consistency or appropriately correlated with conceptually relevant external criterion measures.

This resulted in restructured versions of Clinical scales 1-4 and 6-9 and a demoralisation scale.

Inter-correlations of RC scales and clinical scales

Inter-correlations of the RC scale are lower than for the clinical scale. This makes greater discriminant validity. The scales assess similar, but not identical constructs.

The RC scales are substantially less saturated with demoralisation than are the clinical scales.

Reliability of the RC scale

The RC scales have higher internal consistency values than the clinical scales.

In all samples the RC scales had quite acceptable internal consistency.

The stability of the RC scales over a short period of time is acceptable and in most cases greater than for the clinical scales.

Validity of the RC scales

Most RC scales are measuring characteristics that are similar, but not identical, to their clinical and content scale counterparts.

The RC scales are correlated with external measures.

Most RC scales and their clinical scale counterparts had correlations of similar strength with conceptually relevant external criterion measures. There was evidence for greater discriminant validity for the RC scales than for the clinical scales.

The utility of the RC scales in differential diagnosis has been demonstrated in several studies.

Interpretation of the RC scales

When interpreting any MMPI-2 scales, clinicians first need to assess the validity of test protocols.

Some of the RC scales can add to the interpretation of MMPI-2 protocols. They can clarify the meaning of elevated scores on the clinical scales and generate influences about test takers independently.

The following strategy is suggested for integrating clinical and RC scale information. Consider that for any particular Clinical scale/RC scale pair there are four possibilities: 1) Neither the clinical nor RC scale score is high. No  interpretations should be made 2) both the clinical and RC scale scores are high, inferences about the core construct for the Clinical sale can be made with considerable confidence 3) the clinical scale score is high but the RC scale score not, one should be quite cautious about making inferences about that the test taker has characteristics consistent with the core construct associated with the Clinical scale . It may be a product of demoralisation.4) the clinical scale score is not high and the RC score is, inferences about characteristics related to the core construct are appropriate. The lower clinical scale is likely to result from the absence of demoralisation.

Demoralisation (RCd)

The RCd scale score provides an indication of the overall emotional discomfort and turmoil that a person is experiencing. High scores on this scale are likely also to have high scores on other RC, Clinical and Content scales.

Somatic complains (RC1)

The cardinal feature of high scores on the RC1 scale is somatic preoccupation. Low scores indicate persons who report a sense of physical well-being.

Low positive emotions (RC2)

High scores indicate a person who lacks positive emotions.

Cynicism (RC3)

Clinical scale 3 has two major components: somatic complains and avowal of excessive trust of others.

The RC3 assesses the second component.

Antisocial behaviour (RC4)

RC4 measures antisocial characteristics.

Ideas of persecution (RC6)

RC6 measures persecutory thinking.

Dysfunctional negative emotions (RC7)

RC7 scores the level of negative emotions.

Aberrant experiences (RC8)

RC8 focuses on sensory, perceptual, cognitive and motor disturbances suggestive of psychotic disorders.

Hypomanic activation (RC9)

RC9 scores the level of activation and engagement with the environment.

Personality psychopathology five (PSY-5) scales

The Personality Psychopathology Five (PSY-5) scales were constructed to assess personality traits relevant to both normal functioning and clinical problems. The conceptualization underlying the scales is similar, but not identical, to the five-factor model of personality.

The PSY-5 constructs

Five constructs were the basis for the construction of the PSY-5 scales of the MMPI-2: 1) aggressiveness, focuses on offensive and instrumental aggression 2) psychoticism, disconnection from reality, including unshared beliefs and unusual sensory and perceptual experiences, and feeling alienated and having unrealistic expectations of harm 3) constraint, risk-taking, impulsivity, and ignoring traditional moral beliefs and behaviour 4) negative, emotionality/neuroticism, a predisposition to experience negatively valenced emotions, to focus on problematic features of incoming information, to worry, be self-critical, feel guilty, and concoct worse-case scenarios 5) low positive emotionality/introversion, limited capacity to experience joy and positive engagement and being socially introverted.

Reliability of the PSY-5 scales

The internal consistency of the scale is good.

Validity of the PSY-5 scales

There is a good fit between the hypothesized model underlying the PSY-5 scales and their factor-analytic results.

There is strong support for the construct validity of the PSY-5 scales.

PSY-5 facet scales

Facet scales were developed for each of the PSY-5 scales. But, there is limited reliability and discriminant validity. Their clinical use is not recommended.

Interpretation of the PSY-5 scales

Aggressiveness

Both high and low scores are interpreted.

Psychoticism

Low scores are not interpreted.

Constraint

High and low scores are interpreted.

Negative emotionality/neuroticism

Low scores are not interpreted.

Introversion/low positive emotionality

Both high and low scores are interpreted.