Last month I talked about the five sets of data that are used to arrive at conclusions in psych reports and pointed out that psychological testing is just one of those sources of data.  This month’s newsletter is dedicated to the
Gold Standard of psychological testing aimed at assessing psychopathology, the Minnesota Multiphasic Personality Inventory (MMPI).

Actually, as the use of the plural in the title of this newsletter implies, there is more than one MMPI.  In fact, there are at least four and maybe as many as
six or more, depending on how you count.  What are they?  Are they all equally useful?  What do they tell you?  How do they work?  To keep things relatively simple, I’ll only talk about three MMPI’s.

First, a little history.  Essentially, the MMPI, which was published in 1943, is the most widely researched and used instrument in clinical psychology.  The MMPI is usually the keystone of all clinical psychological test batteries where the major question concerns the presence or absence of DSM-IV-TR psychopathology.  In this regard, the MMPI is able to provide information not only about psychopathology but about the test taker’s basic personality, their credibility, as well as how they are functioning in the world.  Much to its credit, the MMPI has been in clinical use for almost 70 years.  During this time, hundreds of books and thousands of research articles have been published demonstrating that the test is a valid and reliable measure of personality and psychopathology.  It is this body of literature that gives the test its credibility in a medical-legal context.  In its original form the MMPI consists of 566 true-false questions.  Demonstrably valid versions are available in multiple languages.  The test is standardized in that there is a specific set of procedures for its administration and scoring.

The MMPI is especially useful in forensic  evaluations because of its ability to use a variety of validity scales to detect attempts by test-takers to distort their true psychological condition.  In this area, a significant advantage of the MMPI is that the test items are not transparent.  Even psychologically sophisticated individuals are unable to perceive the true intent of the test items.  This makes the MMPI relatively hardy in its ability to detect individuals who are not being frank and honest and may be trying to portray themselves in either an overly favorable or overly unfavorable light.  The lack of transparency in the items is probably best exemplified by a sample question that does not appear on the MMPI but serves as a good illustration:  “I prefer the Queen of Clubs to the Jack of Hearts.”  Quite clearly, when asked to provide a true-false answer, it is not possible for the test-taker to determine the intent of the question.  In fact, it is not unusual to have forensic patients tell me, “That test was confusing.  I couldn’t figure out what you want.”  Exactly!

Once the individual has completed answering the questions they are scored on a multitude of dimensions or scales.  Each scale is made up of a relatively large number of items.  Any given item may appear on a large number of scales.  The interpretation of these scales is empirical in the sense that the scores have been shown by research to be related to certain behavioral characteristics, personality traits and psychopathology.  This is unlike some other tests in psychology where the items have been written to ask about specific behaviors that the test author is trying to measure.  The vast majority of those tests are useless in a medical-legal examination because they are not transparent and have no method for determining who is telling the truth.

After the MMPI is scored, the interpretation is a two step process.  First, one looks at the scores on the validity scales.  As noted above, these scales have been shown to be able to detect if the person is distorting their presentation and trying to portray themselves in either an overly favorable or unfavorable light.  In combination with scores on some other scales it is also possible to determine if they did not understand or even read the questions.  Regardless, once the validity scales indicate that a person has distorted their presentation, or has not answered the questions in an honest and straightforward manner, no further conclusions can be drawn from the MMPI since it is not possible to determine how their distortion has affected the ten basic clinical scales.

The ten clinical scale scores are obtained from groupings of questions that are capable of providing information about a person’s behavior, attitudes, personality traits and/or psychopathology.  In interpreting these clinical scales, there are multiple books and research articles that indicate what they mean.  The method that is universally accepted for the interpretation of the MMPI is to look at the highest two or three clinical scale scores.  Research has convincingly demonstrated that there is a known relationship between the pattern of these scores and DSM-IVTR psychopathology.  For example, an individual who gets their highest two scores on Scales 2 and 8 is said to have a 2-8/8-2 code type, and is likely to be depressed.  Further, if they receive a higher score on Scale 8 than on Scale 2 they are likely to exhibit signs of Schizophrenia in addition to depression.  If you look at the specific questions that make up Scales 2 and 8 you may or may not be able to determine what it is about the questions that measure either depression or Schizophrenia.  In this sense, the MMPI is said to be an empirical test, or a test that has been derived by experimentation, rather than by human reasoning, intuition or, my favorite, “common sense.”

The Minnesota Multiphasic Personality Inventory-2 (MMPI-2) is the 1989 revision of the MMPI.  As the authors of the revision pointed out in the testing manual, the motivation for changing the test was, in part, concern about questions that were labeled as having “sexist wording, outmoded idiomatic expressions, and references to increasingly unfamiliar literary material and recreational activities.”  In fact, a comparison of the MMPI and the MMPI-2 reveals that there were relatively few items that fell into these categories.  Specifically, in constructing the MMPI-2 from the questions that appear on the MMPI, only 13 items from the MMPI were deleted and only 87 were changed to some degree, with many of those questions showing only minor changes.  There was also a net increase of one item as the MMPI-2 has 567 questions as compared to 566 on the MMPI.  Thus, of the 566 items on the MMPI about 15% of the items were changed in some manner and only 2% were deleted!  Or, alternately, 82% of the MMPI was retained.

Another reason that is often cited for creating the revision is that normative data used to interpret the test were collected on a sample of people who were not representative of the 1989 population of the United States.  However, this comment bears little weight as over the years the MMPI has been shown by research to be applicable, valid and reliable for a wide variety of populations from highly diverse socio-economic backgrounds in different parts of the world.

The above discussion notwithstanding, there are some psychologists and psychiatrists who maintain that the MMPI is outdated or obsolete.  However, there is substantial research evidence indicating that the MMPI-2 is not as good an indicator of an individual’s psychological status as the MMPI.  In order to understand what is occurring it is helpful to realize that the MMPI was the subject of extensive research for more than 50 years before the MMPI-2 came on the scene.  During those 50+ years a body of research data was collected that demonstrated that particular code types, like the 2-8/8-2 referred to above, revealed specific details about an individual’s psychological status.  However, this vast body of research data that has been collected on the MMPI has not been shown with any vigor or clarity to be transferable to the MMPI-2.  Most importantly, independent researchers, who are not authors, publishers or distributors of the MMPI or the MMPI-2, have published data in peer reviewed journals and books that indicate that the conclusions about code types determined by the MMPI are not easily or universally applied to the MMPI-2.  In this regard, a person who receives a 2-8/8-2 code type on the MMPI-2 may not have the same psychological characteristics as a person who receives a 2-8/8-2 code type on the MMPI.  Thus, I believe that there are ample data in the psychological research literature that continues to support the conclusion that the use of the MMPI represents a much more conservative approach to psychological assessment than the use of the MMPI-2.

In further support of the conclusion that the MMPI is preferable to the use of the MMPI-2 is the July, 2008 test distributor’s release of what is called the MMPI-2-Restructured Form (MMPI-2-RF).  The MMPI-2-RF is based on what are called the Restructured Scales (RC Scales) that the test distributor, Pearson Assessments in Bloomington, Minnesota, states “provide a more clearly focused interpretation” of the clinical scale scores.  Unfortunately, the use of the MMPI-2-RF has not been met with anything approaching universal approval from the scientific community.  Essentially, the publication of the MMPI-2-RF further confuses the issue of the validity of the MMPI-2 as that publication has only marked the beginning of the needed process of independent research conducted by professionals who are not associated with the publication, distribution, marketing or sale of any of the MMPI’s.

In short, I believe that the publication of the MMPI-2-RF was tantamount to admitting that the MMPI-2 was inadequate and needed help.  I also believe that the ambiguity concerning the usefulness of both the MMPI-2 and the MMPI-2-RF will not be resolved for many years while we wait for the results from independent researchers to clarify the situation.  Thus, until that time comes, if it ever does, the MMPI will maintain its prominence and superiority in the assessment of psychopathology.

* Further information on the MMPI’s, as well as more than 80 of the most frequently used tests in workers’ compensation and personal injury litigation, can be found in my recently published book, Psychological Evaluations in Litigation:  A Practical Guide for Attorneys and Insurance Adjusters.