Yeah, thats definelty a problem with binary classifications per category on Myers Briggs. Percentile based models are better.

Lobotomite

I think the retest issue is mostly that most people have moderate scores on most categories, so if those shift a little bit they get a different score.

For instance, I've gotten INTJ, INTP, and ENTJ at different times because I'm right in the middle of introvert/extrovert and judging/perceiving. However, I get NT every time because I'm way out on the extreme end of those distributions.

bitcoin

charts_and_maps

HealthAndFitness

How Bitcoiners' Myers-Briggs types differ from the general population

k00b

I think the retest issue is mostly that most people have moderate scores on most categories, so if those shift a little bit they get a different score.

For instance, I've gotten INTJ, INTP, and ENTJ at different times because I'm right in the middle of introvert/extrovert and judging/perceiving. However, I get NT every time because I'm way out on the extreme end of those distributions.