The 'Big Five' test of Costa and McCrae, and Eysenck's Personality Inventory, have been evaluated and found to be at least reliable (i.e. give similar results for the same person at different times, and when scored by different people). They are valid (i.e. actually measure some of the things they're supposed to) to some degree: e.g. there are some consistent differences between people high and low on extraversion, or on neuroticism, both on behavioural and some physiological measures. I would not advise using even these tests as the main measure for diagnosing people, or assessing them for jobs, however.

The Myers-Briggs, though frequently used, has not been found to be even particularly reliable, let alone valid. Also, it is based on an outmoded attitude to personality: that it can be classified into distinct 'types'. In fact, the evidence is that personality characteristics vary more according to continuous dimensions ('traits'). For example, extraversion varies continously, with some people being at one extreme or the other, and most being at various points in between, rather than people being either introverts or extraverts.

and sadly it is used/pushed by many major corporations to help justify certain behaviors patterns, ways of treating employees.

is is totally enamored of the Myer-Briggs test. She tends to see much of the world through the filter of that test, and I have pointed out to her that it's neither reliable nor valid, and she just doesn't hear me.

I remember the first time I even heard about it, and it sounded like crap to me.

That is a joke. I am not up to date with modern psychology, but I assume they are mostly for fun.

job selection process. And, for some reason I can never understand (any financial incentives?), tend to use the exceptionally unreliable Myers-Briggs.

My dad was telling me about the psychology tests he had to take every few months when he was in the Air Force in the early '60s, apparently because his main job was loading nukes on planes.

I doubt any of that had a lick of scientific basis to it, yet the Air Force was using these tests to determine mental fitness, or something.

So, I'm calling bullshit.

Essentially, you answer a bunch of questions which boil down to, “Do you prefer A or B,” and the Myers Briggs reports back, “You have (some degree) of preference for A over B.” (That degree of preference varies based on how commonly you choose A over B.)

The mistake that some make is losing track of the idea that it is a continuum from a strong preference for A to a strong preference for B. If you have a weak preference for A over B, you might test one way today, and the other tomorrow.

You cannot assume that you can create 16 boxes, fit everyone neatly into one of those 16 boxes, and assume that they will stay put.


However, it can be quite valuable (in my opinion) to have a sort of accounting to see, “Yes, this is the way I prefer to work.” One of my friends tests out quite similarly to me, while a third tests significantly differently (just as we predicted.) We see it play out in our decision making all of the time.

While the first friend and I are wont to think over a decision, weighing various alternatives, trying to produce an optimal solution, our other friend quickly becomes impatient with this, wishing we would, “Just make a decision!”

... M-B has found exceedingly little acceptance outside corporate bigwigs who would very much like a cheap, effect method of keeping sociopaths, thieves, and liberals out of their candidate pools. Amongst psychologists, M-B has a very poor reputation.

…especially by people who don’t understand it.

“Amongst psychologists, M-B has a very poor reputation.” This is an interesting statement, as I was first introduced to a similar instrument was in a psychology class. On the other hand, years later, I sat in a group where someone from HR used yet another, less specific instrument, and used it rather poorly.


I think this list gets things about right.

http://www.capt.org/mbti-assessment/reliability-validity.htm

While the tests are designed to weed out liars by asking the same questions repeatedly, albeit with minor variation, the tests are still easy to game by anyone capable of simple pattern recognition. This makes validity highly questionable in the best of circumstances.

Also, whether or not personality test results can be used as accurate predictors of behavior is a similarly dubious assertion.