The Minnesota Multiphasic Personality Inventory
, or MMPI, is one of the most-used personality test
s for diagnosing a wide variety of psychiatric
ills, and its method and results especially lend themselves to use in a courtroom
setting as part of expert testimony. The MMPI uses a battery
of anywhere from 399 (abbreviated form) to 600 (full form) true/false question
s to build a personality profile on ten clinical scale
s; there are also three validity scales (six, for MMPI-2; see below) to help detect a concerted effort to bias
the test (though this is more difficult than with simpler tests such as the Thematic Apperception Test
, it is allegedly easier than with the Rorschach
). The MMPI is intended to be administered to American
adults and adolescents
(except in the case of the MMPI-2) with basic literacy
; it should not be administered to children or the mentally handicapped
The clinical scales of the original MMPI are:
Scale 1: Hypochondriasis (Hs)
Scale 2: Depression (D)
Scale 3: Hysteria (Hy)
Scale 4: Psychopathic Deviate (Pd)
Scale 5: Masculinity/Femininity (Mf)
Scale 6: Paranoia (Pa)
Scale 7: Psychasthenia (Pt)
Scale 8: Schizophrenia (Sc)
Scale 9: Hypomania (Ma)
Scale 0: Social Introversion (Si)
These were intended to be pure measures of each disorder, in which a high Sc score, for example, would indicate a high degree of schizophrenic tendency. However, it quickly became apparent that not only was this a simplistic way to interpret results, it was nearly useless compared to the richness of detail that could be obtained by using the profile of the two or three most significant "elevations" in a set of results. Researchers began using just the scale numbers and degree of elevation as a shorthand; as some of the diagnostic categories have fallen out of general usage (such as "psychasthenia"), this tendency has increased. Use of the original scale category names is now more frowned on than not. A 24 profile is a result with 2 and 4 significantly elevated, with 2 being more elevated than 4. Vast amounts of data on the personality tendencies of people with particular result profiles have been amassed.
The validity scales of the original MMPI are L, F and K. The L scale measures the tendency of the test-taker to attempt to portray themselves in a good light. The F scale measures the test-taker's attitude toward the test, and if the results can be considered valid, a high F score indicates a higher degree of pathology generally. The K scale, also known as the defensiveness scale, attempts to measure the same factors as the L scale but in a more subtle manner. Interestingly, takers with more education or who are from higher socioeconomic groups tend to score higher (sometimes dramatically so) on the K scale but lower on the L.
The original MMPI, developed in the 1930s, was criticized because the "norms" on the various scales were set in a statistical fashion from the original personality data used to create the test; since the original sample was Minnesotans (an expected state of affairs; the creators were employed by the University of Minnesota), the test allegedly betrayed a bias such that "normal" was defined in terms of the behaviors, beliefs and attitudes of the average Minnesotan, who might differ dramatically from the "normal" resident of other parts of the US. The questions themselves also were criticized for biases that ranged from quaint to religiously, sexually and otherwise offensive. As a result, the MMPI-2 was developed, and with it the MMPI-A for testing adolescents (the original MMPI was applicable in limited fashion, and with careful interpretation, to adolescents and occasionally even bright pre-teens, but the MMPI-2 is not). Besides addressing the overall norms and the content of the test, the MMPI-2 normalized the statistical "T-scores" of the various components with respect to each other (excluding scales 5 and 0, which are considered to measure different aspects of personality in different ways than the rest); the original MMPI's components varied widely in his respect which made it difficult to compare magnitudes of deviation between components in a statistically valid way.
The MMPI-2, developed in 1989, besides updating the questions and renormalizing the scales based on a national sample, added three new validity scales: FB (F scale, Back side, back meaning the second half of the test), VRIN (Variable Response INconsistency, which attempts to detect a test-taker playing "connect-the-dots" or answering at random), and TRIN (True Response INconsistency, which measures a respondent giving the same answer to mutually-exclusive items). There is also the "?" or Cannot Say "scale" which just measures the number of questions not answered (it's generally agreed that tests with more than 30 omitted items should not be interpreted at all).
Like any other psychological test, the MMPI cannot be interpreted in a vacuum. While "cookbooks" and "blind interpretation" diagnostic software that retrieves profiles from them exist, the history and current life situation of a particular patient may significantly alter the diagnosis. This concern has grown along with the widespread use of the MMPI as an aid in selecting candidates for advancement in government, military and business settings, especially as the computer diagnostic programs have become commonplace and lent an air (perhaps unwarranted) of precision or absolute objectivity to the results. There is no way to make absolute conclusions from the test results alone in the absence of a clinical familiarity with the person tested. Various people also raise issues with what they perceive as the inherent unsuitability of a statistical approach to defining "pathology" in the psychological arena, or to specific methods of gathering or interpreting personality data generally.
Big names in MMPI research and the MMPI testing industry are James Butcher and Alexander Caldwell.
"MMPI", American Academy of Matrimonial Lawyers. A very detailed account of the history, methodology, and current practice of the MMPI, particularly as it relates to legal concerns. http://www.aaml.org/MMPI.htm
"The seriousness of statistical mistakes", Clifford Johnson writing in RISKS Digest 12.35. A contrarian view of the MMPI and personality testing in general. http://catless.ncl.ac.uk/Risks/12.35.html#subj9