Quote Originally Posted by Subteigh View Post
First, a hypothesis (necessarily based on observation) has to be published in a Respectable scientific journal such as the Annual Review of Psychology. Otherwise, there is nothing to disprove.
This. (Although I would say that it's the experiments that are based on observation).

There's nothing to prove or disprove in Socionics or MBTI, because they're just perhaps agreed-upon observations by a couple of people who make up these things. In the same way, you can't "test" Socionics or MBTI because there's nothing to test. A "test" is simply to say "If X were true, then Y would happen". That's basically what a scientific theory is in a nutshell. An experiment would try to set up an environment in a way that Y would occur or not occur, by the condition of X.

Science isn't just a matter of agreed-upon observations; science is about theories and hypotheses.

Socionics "can't proven" not because it's so magical and it's beyond what science is capable of, but rather it's because there's nothing to test, there are no theories and hypotheses.