The reason I poo-poo it is that I see many people using these results as nearly the final word on how to judge someone, rather than just a guide. Which is dangerous, in my opinion. The map isn't dangerous in and of itself, but it becomes dangerous when the mapmaker gives the impression that it gives more detail than it actually does.
edit: Additionally, there's a large psychological bias on the first information you receive about a person. If you see these results before meeting the subject, you will be likely to frame him according to the results, whether they fit him or not.
People are judging each other left, right and centre on much more biased criteria every day whether you like it or not.
Don't have an instrument to judge someone with? They'll judge you based on their preconceptions, their life experience, and a number of other factors that are quite likely more unsound than an external instrument.
People are going to judge you anyway, buddy. Just hope the guy calling the shots isn't an arsehole and has your best interests in mind. Which leads me to believe that your problem isn't with the MBTI itself but the human tendency to judge one another on any basis of criteria. That's life.
Those looking to apply MBTI for any particular reason to any particular person (including themselves) isn't any worse or wrong or ill-meaning than the next fellow. It just depends on whether they're an arsehole or not. Please, don't confuse the two.
The problem here is that these assessments make people think that they have a strong scientific basis for making those judgments. Obviously people always judge each other, but they will feel far more validated in doing so if there's a team of guys in lab coats behind them.
Then the problem is with the people misusing the results, not with the test itself. The MBTI doesn't claim to be a perfect predictor at all; instead it's supposed to measure tendencies.
> The map isn't dangerous in and of itself, but it becomes dangerous when the mapmaker gives the impression that it gives more detail than it actually does.
I agree. I've seen this misused quite a lot.
Psychometric testing is interesting, and can reveal tendencies - but that's about it. Certainly using this as "hard data" for things like hiring, corporate restructures, etc, is fraught with issues.
In reality, the best predictor of a person's behaviour is (a lot of) detail on past performance. Even then, there are flaws.
edit: Additionally, there's a large psychological bias on the first information you receive about a person. If you see these results before meeting the subject, you will be likely to frame him according to the results, whether they fit him or not.