Copyright © 2007-2018 Russ Dewey
Personality tests emerged in the 1930s. Previously, psychological testing had been used to place students at an appropriate level in school (educational testing) or for measuring intelligence.
The new personality tests of the 1930s were aimed at helping therapists determine a client's problems, as a guide to treatment. As a side-benefit, they provided a way for researchers to measure personality traits.
One of the first such tests was the inkblot test or Rorschach. It is still one of the most widely used personality tests.
Hermann Rorschach, when he was a little boy growing up in Germany, enjoyed a game of interpreting inkblots. Children would get a few pieces of paper, fold them down the middle, and put a drop of ink in the middle of each, on the fold.
Then they would flatten the folded paper, spreading the ink inside it. When the paper was unfolded, it contained a symmetrical ink design which often "looked like something" the same way clouds often "look like something," requiring only a bit of imagination.
Where did Rorschach get the idea for his test?
Rorschach and a sampling of inkblots
When Rorschach was a young man, he needed to develop a psychological test, and he remembered the inkblot game of his childhood. He prepared a series of inkblots and had them professionally reproduced.
Contrary to what is widely assumed, Rorschach did not use random inkblots; he prepared them carefully. Then he selected the ones that seemed to him evocative but ambiguous.
He found when people were asked to interpret the inkblots (to say what they looked like) they gave different answers reflecting their state of mind. For example, somebody who said the uppermost of the three inkblots looked like "a monster coming to get me" might reveal fear, while a different person might see an angel, reflecting religiosity.
What is a projective test, and what does the name mean?
This type of test is called a projective test. The person taking the test comes up with an interpretation. The interpretation is a projection of the person's own mental processes onto the picture, which has no pre-assigned meaning.
Because the Rorschach has been used for many decades, clinicians have collected large amounts of data relating common response types to various mental or behavioral disorders. The test can provide useful predictive information (as can any other test administered to thousands of people).
The problem with Rorschach tests is similar to the problem with lie detectors, handwriting analysis, and other tests designed to reveal hidden dimensions of personality. Experts make certain-sounding judgments, based on intuition, but when their judgments are double-checked with other methods, the expert judgments often prove no more accurate than guesswork.
As Paul Meehl argued so persuasively over his 50+ year career, computers using large bodies of data can be useful for making predictions. But when experts add their opinions to correlational analysis, the predictive value of tests goes down, not up.
This issue came up in a psychology department with a two-year clinical Masters program, already under pressure to fit all needed courses into a two-year framework. A new director of clinical studies suggested teaching students to administer the Rorschach, because it was one of the most commonly used clinical tests.
Shortly before, journal articles showed that computerized interpretation of Rorschach responses could outperform human experts, in the classic fashion described by Meehl. Therefore Rorschachs could now be administered by computer.
But the clinical director wanted to add graduate courses on how to interpret the Rorschach. That was necessary, he said, because the interpretation could be highly complex.
Other department members pointed out a contradiction. If the test was objective (like the MMPI, below) complex interpretations would not be necessary. If students had to be taught how to make the interpretations, the test was not objective.
In the end, courses on the Rorschach were not added to the two-year clinical curriculum. Graduate students were taught about the Rorschach in courses on psychological assessment, but they were only taught to administer objective tests such as the MMPI.
The MMPI or Minnesota Multiphasic Personality Inventory started in the 1930s as the Medical and Psychiatric Inventory published by psychologist Starke R. Hathaway and psychiatrist/
The MMPI was designed to make predictions about treatment outcomes of clients at mental health facilities. It was called a personality inventory because it rated an individual on various traits. The traits were chosen for their importance to mental health treatment professionals.
What is the MMPI? How was it developed?
The main advantage of the MMPI was its objectivity. It consists of 567 true/false statements, so it can be administered by anybody and quickly scored by hand or machine. The MMPI is easily scored (or administered) by computer.
Hathaway (quoted in Mednick, Higgins, & Kirschenbaum, 1975, pp. 350-351) said the "real impetus" for developing the test came from reports of insulin shock treatments with schizophrenics. Reports of success of the insulin treatment ranged from zero to 100%.
Clearly the hospitals did not have an effective way to pick patients who might benefit from the treatment. Hathaway saw a need for an objective diagnostic test that would produce reliable results and allow hospitals to decide who would benefit from particular treatments.
The true/false items on the MMPI were initially given to groups of mental patients. They had been diagnosed with well-known psychiatric categories (such as depression or schizophrenia).
The responses of these people were compared to responses of normal people such as visitors to the hospital complex and local workers. Only items that discriminated between mental patients and other people were retained in the MMPI.
How were items selected for the MMPI?
The ten original, clinical scales of the MMPI measured things like hypochondriasis (the tendency to believe one is ill) and paranoia (suspicion of others). The vast amount of data accumulated in the process of refining and validating the MMPI permitted the test to be used for other purposes.
Over the years, over 400 new scales were developed for the MMPI and its successors, the MMPI-A, the MMPI-2, and the MMPI-2-RF. Each scale correlates the answers to questions with different behavior disorders or personality traits.
The original MMPI included three validity scales designed to determine whether the test results could be trusted. One scale (the F scale) detects whether the test is being taken seriously or manipulated in an attempt to sound sick.
This scale is measured by items that are not frequently endorsed by any psychiatric populations, but which indicate the test is not being taken seriously. One example is, "It would be better if almost all laws were thrown away."
Early field-testing showed that nobody who was serious (including psychiatric patients) endorsed such a statement. However, people who were answering at random, or trying to sound crazy, might say True to such a statement.
How did the validity scales work on the original MMPI?
Another scale (the L scale, sometimes referred to as the lie scale) asks about things that would be true of most people, such as "I gossip a little at times." A person who answers False to such a statement might be unwilling to reveal anything negative, because (presumably) everybody gossips a little, so an honest person should be willing to admit that.
A score of 70 or above on any scale is supposed to indicate a problem which might deserve attention. Today the scales are usually referred to by number, rather than by their original names, because the original names may be misleading.
A person who scores high on the schizophrenia scale (scale 8) does not necessarily have schizophrenia. The MMPI would not be needed to diagnose schizophrenia anyway, because the symptoms (such as hallucinations and delusional thinking) are distinctive and well defined.
A high score on scale 8 would simply mean a person answers true/false questions in a way that resembles past schizophrenic patients. That might be useful information to a clinician.
Why are the MMPI scales now referred to by number, rather than their original names?
Other popular personality assessment tools exist. One of the best known is the Myers-Briggs Personality Inventory, which we will discuss in connection with Jung's theory of personality traits, later in this chapter. It overlaps quite a bit with the Big Five list of traits.
Personality tests have been used in many thousands of research projects. However, these tests do not necessarily measure personality as most people use the term.
The previous page on identical twins implied this. It revealed a disconnection between personality tests and what people normally call personality.
Identical twins were said to be eerily similar in personality (individual quirks, sense of humor, etc.). However, large studies of identical twin data by Bouchard and Hays showed identical twins were not unusually similar on personality tests.
Tests like the Rorschach and MMPI were based on intuitions of clinical psychologist and psychiatrists, mostly psychiatrists. Today they are used by clinicians to make assessments (gather information about a new client, before embarking on therapy). The tests were never intended to identify individual quirks, sense of humor, and so forth.
Any large pool of data (lists of questions, observations, web browsing data, patterns observed in the iris of the eye) could be useful in making assessments. Big data allows a computer search for correlations, and correlations (as long as they last into the future) allow predictions.
However, the Rorschach and MMPI are not very good predictors of individual quirks and foibles and preferences. Are there better options? The Big Five traits (and tests based upon them) might be most useful if your goal is to distinguish between people. The Big Five traits were selected for that purpose.
If your goal is to capture what people usually mean by personality, the most straightforward approach is to use the Big Five, supplemented by unique information about a person's habits, quirks, or personal preferences. That should show the differences between personalities most clearly.
Mednick, S. A., Higgins, J., & Kirschenbaum, J. (1975). Psychology: Explorations in behavior and experience. New York: Wiley.
Write to Dr. Dewey at firstname.lastname@example.org.
Don't see what you need? Psych Web has over 1,000 pages, so it may be elsewhere on the site. Do a site-specific Google search using the box below.