So combining two biased samples balances out? Or is it more testing self perception? E.g. do you think of yourself as X not whether you are objectively X. Probably thinking about this question in detail ruins its usefulness as a text question anyway....
Offhand I don’t entirely know, but the tests do work so your comparison issues can’t be serious. If they’ve been normed for your group, perhaps that screens off the worst of the issues.
I’m only talking about myers briggs here, but I’m really confused about why you say the tests seem to work.
One of the main criticisms of MBTI is actually its lack of reliability—that is, its lack of consistency from test to test. MBTI in general is rubbish and when I talk about personality tests or correlations, it’s always Big Five or some other decent test.
Don’t worry; the people whose responses the tests were built on didn’t have access to a random sample either! Still seems to work, though.
So combining two biased samples balances out? Or is it more testing self perception? E.g. do you think of yourself as X not whether you are objectively X. Probably thinking about this question in detail ruins its usefulness as a text question anyway....
Offhand I don’t entirely know, but the tests do work so your comparison issues can’t be serious. If they’ve been normed for your group, perhaps that screens off the worst of the issues.
One of the main criticisms of MBTI is actually its lack of reliability—that is, its lack of consistency from test to test. MBTI in general is rubbish and when I talk about personality tests or correlations, it’s always Big Five or some other decent test.