Here’s a correlation of Myers-Briggs results and personality disorder among college students:

http://www.uccs.edu/Documents/dsegal...ures-JPT-2.pdf

The conclusion: people with I, N, T, and P on the Myers-Briggs were more likely to correlate with personality disorders.

Another way to look at it: If you start from Jung’s theory of innate preferences, then people with I and N preferences are in the minority of the population, in the US at any rate. And people with T preferences aren’t as “friendly” as people with F preferences.

So maybe there is an inherent bias in the majority, people-oriented population to consider people with less common Myers-Briggs preferences as being "abnormal"? And hence to develop ways to measure personality disorder that confirm that?

When I was in graduate school 20 years ago I took some psychology statistics classes with a man who was a big fan of the Myers-Briggs. He thought that starting from the basic Myers-Briggs preferences different people would tend to develop different personality disorders, if other risk factors were there (i.e., parenting and environment)

I agree with Panda, personality type or disorder (if any) is different from identity, which is who someone is as person, just who they are. Sometimes who we really are is hard for us to know, especially during adolescence or times of stress or transition and the tests seem to offer some insight. I wouldn't sweat it, though, if it's hurting and not helping.