The Problem with Standardized Personality Measurement Tools
They are devoid of context. Utterly devoid of it. I routinely run tests on people for their conserved personality traits. I can't tell you how, because the methods differ depending on the situation. All I can tell you is that the personality traits are indeed conserved.
One thing I do do, is test the same trait in a wide variety of circumstances. I do this by first assuming the person's mentality, and then asking how they would view the current situation. I then reach out to that person in the context of the assumed mentality. If they respond in the affirmative, then the response may be a sign of a conserved trait. Especially if they hold to those guns under ensuing criticism. The criticism creates a stream of consecutive tests which affirm that the personality is holding up under pressure, as I believed it would. One single test sets up a dynamic of successive further proofs of the initial hypothesis.
If we're to successfully make a fool-proof test, then we must communicate heart to heart with the person who is being tested. The questions themselves must reach to the heart of the person and speak to their point of view. The person undergoing the testing selects from those points of view those which, in their estimation, are most similar to their most comfortable personal standpoint. After making their selection, an effective test will, in fact, issue a string of attacks upon that standpoint. It then becomes a question of whether or not, in the face of these attacks, the individual will back down. An interactive test would allow the individual to take back a standpoint which they felt did not withstand critique. A person has the type they have because they can pull up, confidently, infinitely many reasons for holding to the point of view associated with that type. We don't need to test each of the reasons a person may hold to a type, but we can ascertain, via a pointed inquiry as to whether or not they believe their position indefinitely valid, whether or not they have a counter-position ready to the argument. Typing politics may prove instrumental in deducing point of view; however, we need not factor in the political dimension to determine type: pointed reference to specific political manifestations of functions use alone suffices for this task.