I had femininity undefined. I guess they only define masculinity if you check the "male" box at the start of the test and vice versa.
The higher the T numbers, the more a mental problem is supposed to apply to you.
I was honest at answering the test, but got a bit sloppy after a while and did not look up each single word I didn't know and for the questions I did not understand, I clicked randomly. Moreover, I probably overlooked some of the "not"'s in some questions and so did not always get the double negation questions right ("I do not..." - False/Right), maybe because I'm no native speaker. Test result for me therefore was that my whole test result is unreliable and I might be cheating (I looked it up in a German page).
Otherwise, the results were relatively accurate.
The basic idea behind the MMPI-2 test is good, it's not based on any psychological theory, but on statistic findings they had with people who presented with specific problems/diagnoses and tended to show according patterns when answering the test. However, the test result may be a bit unreliable or biased as they might have misdiagnosed some of those patients, plus some of the questions are somewhat dependent on one's culture, plus in some cultures, more people tend to be sloppy when filling in very long tests than in others. And because e. g. if your health is really bad and you answer the according physical health questions correctly, you may get scored as a hypochondriac (a result I did not get, by the way, my health is good).
I guess this test is less reliable if you do it when you have trouble concentrating or if you are no native speaker. So if you get asked to make this test, warn the shrink if it's not your day, and if you're no native speaker, keep him present to ask about vocab stuff and take more time so you don't overread the "not"s and "don't"s. Some questions are asked twice, once in a positive way and once with "not", I think these are the ones which are supposed to catch you if you're cheating. In my opinion, this test is a good screening method to find possible problems which you can discuss with your client, but it's not reliable enough to use the results as a "proof" of whatever, or in an exam-like situation such as a job application or trying to get the okay for hormones or surgery.
