The Minnesota Multiphasic Personality Inventory (MMPI) one of the most popular personality tests today is often the standard to which others are compared (Cohen & Swerdlik, 2010).
It was designed to aid in the diagnosis and prediction of mental illness in patients 14 years of age and older, in a more objective manner. The MMPI is atheoretical and therefore was not created based on a specific theoretical background.
It initially contained 566 true and false questions divided into 10 clinical scales, which were derived through research of various sources including personality items that had been previously published by others, textbooks and psychiatric case studies and reports (Cohen & Swerdlik, 2010). The construction of the test relied heavily on empirically derived items. The test developers presented the scale items to two groups, the clinical criterion group which consisted of members of a population who were believed to have the same diagnostic condition, mostly psychiatric inpatients, and a normal control group who were non-diagnosed individuals.
This group is also known as the standardized sample and included 1500 people from a variety of backgrounds and locations.
After the items were administered to these groups any items that differentiated the two were kept for further consideration. It was from these items that the scales of the MMPI were created (Cohen & Swerdlik, 2010). Over time the developers became aware of research regarding problems with self-report methods and developed three validity scales which would indicate deviant responses that could affect the test results.
These scales are known as the L (Lie) scale includes 15 items in which people may attempt to portray themselves in a more favourable light. The F (Frequency scale) includes 64 items that would not normally be endorsed by normal individuals. The K (Correction) scale also attempts to indicate portrayal of oneself in a more favourable way but in a more subtle manner, a high score indicates defensiveness. Another scale that was included was the “Cannot Say” scale which measures the number of items that the test taker did not mark a response to or marked “cannot say”.
If this applies to 30 or more questions the test is no longer considered valid, some indicate that a mark of 10 on this scale is cause for concern. The MMPI consists of 550 true or false questions with some versions including 16 repeated questions, totalling 566 questions. Scores are in the form of T scores in which the mean is 50 and the standard deviation is 10. The MMPI also includes content scales, their purpose is to group similar test items together to give the impression that all items in the scale are related.
Hundreds of supplementary scales have also been devised since the original test was published, which focus on a variety of areas (Cohen & Swerdlik, 2010). Initially, the MMPI was administered as a paper and pencil test, but is now available online, on disk for computer-based use, on index cards and even in an audio format. The audio format is designed for use with individuals who have difficulty reading and are instructed to vocally answer the questions. For administration of all other versions it is imperative that the test takers have at least a grade six reading ability.
There is no time limit (Cohen & Swerdlik, 2010). It is no longer necessary for people to manually score the test; a computer scoring service is now available and is the method most users choose. Analyses provided can range from a numerical or graphical representation or may include detailed descriptions of the results. As for interpretation of the results, they were not meant to be based on single scores but on the patterns and profiles that emerged from all of the results together, however this proved to be extremely difficult.
Since then two systems have been developed to interpret the scores, the Meehl system, which consisted of compiling the data into a possible 40 different codes types which each had a specific meaning, and the Welsh code which used a formula to analyze the test takers responses, then further interpreted the scores (Cohen & Swerdlik, 2010). More recently the MMPI-2 was created. The most significant difference between it and the original is that a more representative standardization sample was used.
It remains very similar to the original except that approximately 14% of the items were changed to reflect the change in attitudes of the general public. It contains 567 items, with 394 that are original, 66 revamped and 104 entirely new questions. It also includes an Fb scale which was included to identify questions in which the test taker is no longer diligent upon answering the questions accurately and consistently. However, criticism arose again, and the MMPI-2 was in need of revision. Tellegan et al. (2003 as cited in Cohen & Swerdlik, 2010) stated that there were too many overlapping items in the MMPI-2.
The lack of discriminant validity and individuality of the scales could lead to confusion upon interpretation of scales with higher scores. In devising the MMPI-2 Restructured Form Tellegan et al. (2003 as cited in Cohen & Swerdlik, 2010) maintained the core components of the MMPI-2 scales but created a revised measurement system. They restructured the scales using the original items but removed the demoralization factor from them and instead created a new demoralization scale. They also used the normative sample from the MMPI-2 but combined it with data from their own research of three different samples.
Further, they included more scales that measured higher-order functioning and another validity scale to measure infrequent somatic response. They stated that these modifications had improved both the convergent and discriminany validity of the MMPI-2 and created a psychometrically sound assessement. This version contains 338 items and 50 scales. The MMPI-2 RF has also been formatted into a version for use with adolescents between 14 and 18 years old, known as the MMPI-A. The objective of the MMPI and its revisions is to assess pychopatholgy in adults, while the MMPI-A is intended to measure psychopatholgy in adolescents.
It can be administered via internet, computer disk, audio tape or paper and pencil. Scoring may also be done via computer or manually, however the later is rarely used. The MMPI has undergone revisions that improve its validity and usefullness in the 21st century. Due to the large number of items and scales it is used widely in both clinical and nonclinical settings to assess the psychopathology of individuals as well as to predict the possibility of certain behaviours occuring. Sellbom and Bagby (2010) experimented with the use of the validity scales of the MMPI-2 RF to detect overreported psychopatholgy.
Due to the inclusion of a new validity scale, as well as the revision of the validity scales from the MMPI-2, the researchers stated that it was necessary to investigate whether the cut scores continue to be valid across situations and settings. They also wanted to determine if coaching the participants had an effect on the validity scales. Because of the amount of information readily available to the general public they felt this investigation was necessary. The researchers used an anologue simuation design and recruited 219 University of Toronto students and 146 inpatients with severe mental disorders for the tudy.
The psychiatric patients completed the MMPI-2 RF as part of a routine psychological evaluation, while the participants took the test at two separate times. The first time they filled it out as they normally would, then after a break they took the test again and pretended to have a mental illness. One group of participants was coached regarding the validity scales, how they worked and how to avoid being detected. They found that the Infrequent Psychopathology Responses scale was best at determining faking with a large to very large effect size, regardless of whether the individuals had been coached or not.
In another study Michael et al. (2009) used the MMPI-2 to investigate the potential outcome of psychotherapy in a clinical setting. The researchers had each of the 51 patients undergo a psychological assessment which included the MMPI-2 as well as an interview and an Outcome Questionnaire-45. Those who had a OQ-45 score of greater than 63 were considered eligible for the study due to their increased level of distress. All received a diagnosis on Axis I or II of the DSM-IV and received a combination of cognitive-behaviour therapy and interpersonal therapy.
At each subsequent visit the patients completed a OQ-45. The initial OQ-45 taken at intake, as well as those on all further visits were used in the analysis. In this case, they found that a higher score on the Hypochondriasis scale, the Depression scale and the Hysteria scale indicated poorer symptom reduction as shown by the OQ-45. The previous two studies were examples of how the MMPI-2 can be used in different situations. In the first it was used to detect overreported psychopathology, while in the second it indicated whether the patient would have a poor prognosis for treatment.
Since its revision the MMPI-2 RF has been used in a multitude of studies in which the validity of the measure has been evaluated. Researchers have reported that the MMPI-2 RF is psychometrically sound and that it now has fewer item intercorrelations or overlap (Cohen & Swerdlik, 2010). They have also supported the finding that the revisions have increased both the convergent validity (whether items that should be related actually are related) and the discriminant validity (whether items that are supposed to unrelated actually are unrelated) (Cohen & Swerdlik, 2010).
Thus, based on much of the research so far the MMPI-2 RF is a valid measure and accurately measures what it is intended to measure. Due to the fact that so many studies have verified the soundness of the MMPI-2 RF it is fair to say that it is also a reliable measure. It has test-retest reliability, because if the test were to be administered to the same person, under the same conditions, the result would end up the same each time, although the MMPI-2 RF is still relatively new and so independent research has yet to confirm these findings.
It is also necessary to further investigate its use cross-culturally since the MMPI-2 may have been lacking in this regard, one hopes that due to the inclusion of a more diverse normal standardization sample the MMPI-2 RF will overcome the challenges of its predecessor. The population continues to dramatically change with each passing decade and therefore so must the MMPI if it is to remain a valid and reliable assessment tool, as has been demonstrated through the previous revisions. The creation of the MMPI-2 RF has only recently allowed this test to adequately assess psychopathology in the 21st century.