Quote Originally Posted by Singu View Post
We should be looking for TRUE LIEs.

LIEs being rare would increase (or decrease) the probabilistic likelihood of one being an LIE. And we could sort of infer the statistical validity by looking the total ratio of LIEs in this community.

I think this sort of thing should be calculated probabilistically, anyway.

It's all BAYESIAN.
This community is not a representative sample of the whole population, also sample size is very likely too small.

Anyway, I'd like to see a document where LIEs (in socionics) are estimated as especially rare, because I don't believe it's true.