Friday, March 3, 2017

--- Need for novelty ---

The Minnesota Multiphasic Personality Inventory (MMPI) is a 567 item test. Unlike many personality tests, it's not designed to tell you what you're like (since you probably don't already know that). It's designed to tell you what's wrong with you. It's a diagnostic test.

I've taken this thing five times in my life - twice as part of job interviews and three times to fine tune the software I wrote for the evaluation department I was over. As a vocational evaluator, I was not legally allowed to use the MMPI without oversight by a licenced psychiatrist. Tests are controlled like drugs. There were a lot of tests I could give, but not the MMPI. But I would get MMPI profiles in client files and have to incorporate them with the evaluation reports I sent back to the client's counselor.

I like taking tests but the MMPI just hurt my head.

When the MMPI was originally published, it included a scale called the Masculinity/Femininity Scale. This scale was originally intended to diagnose for homosexuality, which was considered a disorder when the test first came out. There was a problem though. The more it was used (It's still one of the most popular psychological tests in use.), the more people noticed that, as men aged, they became more feminine (according to the test). In fact, the scale did not measure what it was originally intended to measure. It measured the social stereotype associated with femininity and masculinity. The Wikipedia article ( accessed 3/3/17) says that the model for masculinity was the Marlboro Man and that for fimininity was Donna Reed or June Cleaver. Scary, that.

The scales have been restructured.

So, what does that mean? Is the MMPI a bad test? Well, not really, It justs means that people that give tests should know the tests intimately. For instance, they should know what the tests they use measure.

A lot of folks seem to think that tests are created by thinking up a bunch of questions to ask that would seem to measure what the test is supposed to measure and put them on paper. If only it were that easy. I designed some inhouse tests for my evaluation department and a lot of work went into them.

To create a psychological test, you start by doing what I said in the last paragraph - you come up with lots of questions that you can ask people and that should tell you what you want to know, but you don't stop there. Next, you give the test to a lot of people (the second edition of the MMPI used 2,600 people for the study group.) The statistics you get tell you things like "Which test items tell you anything useful?" "What do the test items seem to be measuring?" "How strongly do the items measure what they seem to measure?"

After you get rid of the test items that do not distinguish between people (If everyone answers an item the same way, it's pretty useless in a test) or only measure what you want to measure weakly, you give the test to a lot more people. Items that correlate strongly with each other form a scale. You can usually look at the items in a particular scale and see what they're measuring (but sometimes they throw you a curve ball like the MMPI Masculinity/Femininity scale). Then you compare the scores you get from the test with scores you get from other measures of the same things your test is supposed to measure. For instance, if you want to measure mathematics ability, you can compare the scores of your test from the students in a mathematics class with their grades. If all the numbers look okay, then you can publish your test - but it's not over yet.

Sometimes, people who use your test after it's published will notice some problems. Then, you go back to the drawing board, work out the problems, and publish a second edition. The MMPI is in it's third edition, the MMPI-2-RF (Restructured Form). Probably the most popular psychological test ever created, the Wechsler Adult Intelligence Scale is currently in it's fourth incarnation, the WAIS-IV.

If you are taking a psychological test, you are, in a sense, competing with the many test-takers who took the test in the design phase of the test. A test that is supposed to measure manual dexterity will pit you against an expert fabricator. It actually (given that glitches like that in the original MMPI didn't get into the design) does measure your ability to handle little bitty objects. And most tests given by professionals are good tools for measuring what they're supposed to measure.

But most tests that are given by professionals are only supposed to be given by professionals, so you can't just go out and take the MMPI. There are some fun tests that are published for general use, though. Some of my favorite books containing tests are:

The Mind Test by Rita Aero and Elliot Weiner (1981) Harper Perennial
The Brain Game by Rita Aero and Elliot Weiner (1983) Harper Perennial
Know Your Own Mind by James Greene and David Lewis (1984) MacMillian Publishing Company
Who Do You Think You Are? by Keith Harary and Eileen Donahue (1994) Harper San Francisco
The Luscher Color Test by Max Luscher (1990) Pocket

Most of these books can be obtained for next to nothing online since they're "old", but they're still available the last time I checked. I included the Luscher simply because it is so much fun.

The Mind Test has a scale in it called the Desire for Novelty Scale. Just by the name, you'd think that adventurous people would score pretty high, right? So I took it and scored 4, which is low. Was it wrong?

No, the test is designed to identify people who have little novelty in their lives and are unsatisfied with that. I have plenty of adventure.

No comments: