[Question] How good is a human’s gut judgement at guessing someone’s IQ?

In her lat­est post, Sarah Con­stantin writes:

What­ever abil­ity IQ tests and math tests mea­sure, I be­lieve that lack­ing that abil­ity doesn’t have any effect on one’s abil­ity to make a good so­cial im­pres­sion or even to “seem smart” in con­ver­sa­tion.

I re­mem­ber that I made the op­po­site pre­dic­tion a few months ago, while I was study­ing psy­cho­met­rics more ac­tively. In par­tic­u­lar, I had the sense that be­cause of the pos­i­tive man­i­fold (i.e. al­most all pos­i­tive men­tal at­tributes quite strongly cor­re­late), it seems like judg­ing some­one’s IQ should be rel­a­tively easy, and also quite valuable in as­sess­ing a large num­ber of other pos­i­tive qual­ities, which should make it quite evolu­tion­ar­ily ad­van­ta­geous to be able to as­sess it.

Two con­crete ex­per­i­ments I’ve thought about:

1. Take a group of peo­ple whose IQ you now, then just take pic­tures of their faces (or short videos of them say­ing some­thing), and then have a group of par­ti­ci­pants rate them on their in­tel­li­gence (prob­a­bly in or­der). See how strong they cor­re­late.

2. Throw a bunch of par­ti­ci­pants into a group, have them talk to each other, then af­ter­wards ask them about the rel­a­tive judge­ment of the in­tel­li­gence of the other mem­bers of the group, see how pre­dic­tive they are.

Th­ese seem like very straight­for­ward ex­per­i­ments, that I can imag­ine some­one hav­ing already run, and where I would be pretty in­ter­ested in the re­sults. If only so that I can be bet­ter cal­ibrated on how much to trust my gut judge­ment of some­one’s in­tel­li­gence.

Does any­one know of any similar ex­per­i­ments that have been run?