How does this solve the problem of confirmation bias and the fact that MBTI relies on self-report? Suppose most people don't know how to describe themselves accurately. I've seen clear ISFJs test INTP and cling religiously to that description. The test needs to be accurate before it can be used for any real research, which is impossible given that it relies on self-report.
If that sentence can be true then there's more than one test available, one being MBTI itself, the other being the perception (and classification) of the test subject by other individuals moderately well-versed in Jungian functional analysis, or at least, individuals moderately well-versed in type-spotting.
This is the puzzle to me, that people can observe that MBTI
gets it wrong.
So... there was something to get right?
Thus, the important weakness of "MBTI"... that Jungian functional analysis--the theory that the Meyer Briggs Type Indicator was developed to operationalise--doesn't come with a depth gauge. It does provide a model but doesn't describe testable links between the model and reality. In other words, we can't--neither via MBTI nor Jungian function analysis--say something like "Fe is just short hand for the [something that admits scientific testing, such as say, brain activity, or whatever]."
I have the same general problem with, say, quantum mechanics. Quarks? Yeah, right.
(Observe, ladies and gentlemen, the above comment is naive, and uttered knowingly naively. It does not imply anything about quantum mechanics. It does imply something about my understanding of quantum mechanics.)
(Observe too, ladies and gentlemen, the above parenthetical comment is *not* a proof that all persons who disbelieve "MBTI" merely don't understand. It is entirely possible that it is I who doesn't understand.)
It's a model, a categorisation. One that hasn't described its basis for making that categorisation. The evil mastermind left it to us to observe for ourselves. Kinda cultish that way, one does have to say.