Personality Types in Science

[ysabetwordsmith]

A new study has identified clusters of four personality types.  That doesn't mean these are the only ones, just that they're common enough to show up as clusters.  I'm not a close match for any of them.

I have seen some other studies of personality types.  I've looked at all kinds of quizzes.  Most don't fit me at all.  Astrology is embarassingly accurate.  ColorCode is quite accurate, if you throw out the idiotic idea that only the highest-scoring color counts and instead treat them as percentages.  

I think, if people are going to use any personality test for more than idle entertainment, it should be tested for accuracy.  This is especially important given that corporations have gotten tired of using real interviews for job selection and are resorting to personality tests as a hiring filter.  If the tests are bogus, then people are being denied a livelihood on spurious grounds, which is not a good thing at all.

Comments

[curiosity]

My job used one of those personality test hiring filters. Actually, a couple of my jobs have. But the most recent one realized it was complete bullshit and tossed it out. They've gone back to good ol' one on one interviews and gut checks.

Having been required to conduct a few interviews using one myself (though I ignored the results) I can tell you it was ridiculously over-simplified and largely irrelevant to any generation after the Baby Boomers.

Well ...

[ysabetwordsmith]

I think that a well-designed personality test can have its uses. If there's a job that requires certain traits, like resilience or extroversion, and you have a ton of applicants, then weeding out some of the least appropriate makes sense. Or if you hire in batch lots and then sort into departments later, it could help you identify which people would do better in what roles. They're often very illuminating in figuring out why two people are butting heads. But they're not designed for what a lot of companies are trying to do with them.

[curiosity]

Also. I tried to take that ColorCode test but it makes no allowances for the type of childhood I had so I can't answer the questions.

Alas!

[ysabetwordsmith]

That sucks. :( It's a problem I frequently run into with all kinds of tests: the questions are so far off-base that no matter what I answer, it will be wildly inaccurate. If more than a few questions are like that, there's no point in taking the test -- and that's usually what happens for me. The tests are designed for humans, what can I say.

Issues with the ColorCode test

[ng_moonmoth]

Yeah, I found some of the issues you were mentioning crop up in the ColorCode test. There were too many questions that weren't anything like me.

The real deal-breakers on the test came at the end, where they were soliciting demographic information. They wanted a bunch of things I consider personal information. I would not want to reveal them truthfully, am not interested in phonying up the demographics and creating a burner email account just so I could see the results, and don't really care to fuzz whatever data analysis they do that way. And to top it all off, they had a gender category that didn't have a place for me. That really tore it.

Re: Issues with the ColorCode test

[ysabetwordsmith]

Probably because it's an old test. I got it from a book that came out years ago, before gender diversity really started to catch on. :/ Sorry about that.

[gatheringrivers]

MOST of these tests don't test for the one thing I've had to develop my OWN radar for, due to a lot of unfortunate personal experience: Toxic personalities.

Workplaces really need a toxic-person-filter system of some sort, AND the chutzpah to actually USE the thing.

Yes ...

[ysabetwordsmith]

I agree that would be useful, if it could be made to work. However, toxic people tend to lie a lot. That makes it challenging to filter them with test questions.

[mama_kestrel]

Years ago, when Meyers-Briggs first came out, my father in law brought it home one Thanksgiving to test on the assembled family, figuring we were a large enough group to make a decent test pool where he could compare the statistical result against what he actually knew of his test subjects. (I pointed out that we were, one and all, very well educated and articulate, but no test group is perfect.) Between the folks, kids and spouses, there were 10 of us.

Eight got reasonably accurate results. One kid split even on one trait-pairing, requiring knowledge of the individual to know which way to break the tie. And then there was me. I broke the test completely, splitting exactly evenly on 3 out of 4 trait-pairings, with only a 1 point difference in score on the last pair. They were reduced to reading through all the descriptions and finally saying "That one. Has to be that one." (It amused me that the one they chose, with no input from me, said "You are among the approximately 5% of the population that may be genuinely psychic.)

My father in law was trying the test out for usefulness as hiring manager for a huge law firm. He decided, based on the family experience, that if you needed to know the person ahead of time to interpret the test accurately it wasn't useful for pre-screening purposes. He considered the test to be 80% accurate at best, with willing participants who had no reason to try to skew the results. I wish more H.R. departments understood that.

Yes ...

[ysabetwordsmith]

That's probably typical.

I suspect that most people believe in personality tests because they want those things to work. They have more applicants than they can interview, for instance, or they have problems with bad employees.

[paynesgrey]

Hrmm, I don't seem to fit into those four personalities either. I wonder what a test would put me in.
I think personality tests for a job is a joke. I remember if I was job searching and had to take one of those things, I'd give up on the job. Taking a test isn't a proper display of my skillset and personality.