Robert M. Gordon, Ph.D. ABPP

Validity Scales

Definitions of MMPI/MMPI-2 scales: Scales of Validity and Bias


Measures of Completeness and Consistency

? Scale -  Cannot Say (MMPI/MMPI-2). A tally of omitted items.  High scores may be due to obsessiveness, defensiveness, difficulty in reading, confusion, hostility, or paranoia. It is important to look for a pattern that may exist in the items that are left blank. More than ten left unanswered may be of clinical significance. Twenty or more left unanswered should be considered significant.

TRIN - True Response Inconsistency scale (Butcher, Dahlstrom, Graham & Tellegen 1989). (MMPI-2) 23 pairs of items that are semantically inconsistent. High scores, 13 or more, have responded to the MMPI in a "yea-saying" test set, responding mostly "True".  This can be also be High due to severe psychopathology. Low scores of 5 or less responded with a "nay-saying" test set, responding mainly  "False." 

VRIN - Variable Response Inconsistency scale (Butcher, Dahlstrom, Graham & Tellegen, 1989). (MMPI-2) 49 pairs of semantically inconsistent responses. This is possibly the best measure of random, or inconsistent responding. Scores of 13 or more would be considered invalid. This can also be High due to inconsistencies within the personality.

Self-Favorable Scales

L  Lie Scale (Hathaway and McKinley, 1951) (15 items). High: (> Raw 5). Tendency to create a favorable impression as a response bias, conventional, rigid, moralistic, repression, denial, and insightless. A high L can mean anything from a very well mannered normal wanting to give a good impression, to a compensated paranoid. A high L will submerge scales of obvious psychopathology, and inflate scales of healthy functioning such as the Ego Strength scale.  Interview the person to see if they can truly walk on water.  Low: (< Raw 3). Admitting to minor faults and shortcomings, independent, self-reliant.

K Scale - Defensiveness (McKinley, Hathaway & Meehl, 1948). (30 items) K is a subtle and valuable correction for defensiveness. K assumes psychopathology.  If someone with a history of psychological problems scores high, then they are being defensive. However, a high K is also associated with high education and socio-economic status. That is, people who are highly educated and getting along well with other individuals, should score moderately high on the K scale. The K scale was derived from individuals who were hospitalized, clearly having serious psychological problems and yet producing normal profiles. They were being defensive by claiming that they had no psychological problems. High (>Raw 22). If there are signs of psychopathology in the history, then high K indicates defensiveness, insightlessness, intolerance, dogmatism, and being controlling. Very high scores are usually a sign of defensiveness. High scores are common with individuals who are well adjusted and well educated, and tend to be in control of their lives. Low (<T46). Guarded prognosis for any insight therapy since their ego strength is low; masochistic confessors, poor self-concept, distrustful, and angry. A very low K could often be the only indication of psychopathology on an MMPI profile. 

S  Superlative Self-Presentation Scale (Butcher and Han, 1995) (50 items); based on comparing the responses of male airline pilot applicants with the male MMPI-2 normative sample.  Five subscales are; Belief in Human Goodness, Serenity, Contentment with Life, Patience and Denial of Irritability and Anger, and Denial of Moral Flaws.  Highly correlated with the K scale.  As with K, if the person is indeed high functioning, a high score accurately measures ego strength.  If however, the person’s history does not support claims of superior adjustment, and T>65, consider a faking to look good bias.

Mp- Positive Malingering Scale (Cofer et al. ,1949) (26 items).  Developed by having college students take the MMPI under conditions, of fake good, fake bad and normal.  It may measure a conscious attempt to give a favorable impression.  Mp is highly correlated with Sd, L and S.  Mp should only be used along with L and K.  If T greater than 60, and L and/or K are also high, there is likely to be a conscious attempt to fake to look good.

Sd-Social Desirability Scale (Wiggins, 1959) (33 items); based on asking college students to respond to the MMPI as a person who has the general values of the American culture.  Sd highly correlates with Mp and L, and only slightly with K.  High scores, >T65, are associated with claims of assertiveness, confidence, and virtuousness. 

Self-Unfavorable Scales

F Scale - Infrequency (Hathaway & McKinley, 1951) (60 items).  Very high (>T99) possible random, exaggerated, or mis-scored profile. Very high scores (T> 90) commonly found with psychotic patients. High scores (>T70), best measure of overall psychopathology, resentment, acting out, moodiness. Mostly elevations in the F scale are due to psychopathology; high item overlap with scale 8. Low scores (T<45), possible fake good profile.

Fb - Back F scale (Butcher, Dahlstrom, Grahm & Tellegen, 1989). (MMPI-2) 40 items found after item 280. Developed like the F scale, it is made up of items that are endorsed less than 10% of the time by normals (but frequently by disturbed individuals). If Fb is above T99, and F is not high, then the individual may have randomly responded to the latter part of the test. This is more likely than the other possible interpretation, namely that the testee decompensated toward the end of the test from having taken such a long self-report.

Ds Dissimulation Scale (Gough,1954); (58 items);  developed on true neurotics verses normals  faking neurosis, Ds measures more exaggeration of neurotic symptoms, whereas F and Fb assesses more exaggeration of psychotic or severe symptoms.  Scores greater than T65 suggest some exaggeration, and scores greater than T98 are most likely to be exaggerated.

F(p)- Infrequency-Psychopathology Scale (Aribisi and Ben-Porath,1995a, 1995b); (27 items);  the F scale was developed on normals who responded to items less than 10% of the time.  The F(p) scale was constructed of items endorsed by 20% or less by two separate groups of psychiatric patients vs. the MMPI-2 sample.  The F(p) is better than F or Fb in detecting feigning serious psychopathology.  Scores between T71 and T113 may be exaggerated, unless the psychopathology is obviously severe; scores greater than T113 suggest exaggeration.

Go to Basic Scales