Ugh, this study selected its "superior iq" group by surveying Mensa members. Study (https://www.sciencedirect.com/science/article/pii/S016028961...) makes no mention of this selection bias, though even the article does.
Incidentally, Hambrick's writeup here is also quite bad. Aside from failing to mention all of those reasons why it's BS, he doesn't describe accurately the research he does mention. Consider this paragraph from the end:
> All the same, Karpinski and her colleagues’ findings set the stage for research that promises to shed new light on the link between intelligence and health. One possibility is that associations between intelligence and health outcomes reflect pleiotropy, which occurs when a gene influences seemingly unrelated traits. There is already some evidence to suggest that this is the case. In a 2015 study, Rosalind Arden and her colleagues concluded that the association between IQ and longevity is mostly explained by genetic factors.
How did he miss the fact that the Arden study in question shows, as all such studies show, that intelligence correlates with greater lifespan/longevity when he's written an entire credulous column about how intelligence is bad for you and makes people crazy and increases the rate of mental disorders (some of which, like schizophrenia, reduce your life expectancy by decades)? It's right there in the abstract, it's not hidden away. One would think that would at least merit a brief pause to puzzle over the discrepancy...
More seriously, Mensa membership selects for people whose main accomplishment in life is doing well on an IQ test. So they are at least somewhat smart. But if you look for people who have actual accomplishments, you'll find people who tend to also be smart but are more interesting to be around.
Note, this is not sour grapes. My test scores mean that I'm qualified for Mensa, but I've never been impressed enough by any Mensa member that I met to be interested.
1. to identify and to foster human intelligence for the benefit of humanity;
2. to encourage research into the nature, characteristics, and uses of intelligence;
3. and to provide a stimulating intellectual and social environment for its members.
It's not entirely clear to me how anything they do benefits humanity (1), the "nature, characteristics, and uses of intelligence" seem to revolve around games for them which is a pretty meaningless measure of any form of intelligence I care about (2) and while I'm not a member so I can't say whether it's stimulating to them, I can certainly say that all the materials I've come across from Mensa have soon bored me(3). So from where I sit, Mensa fails on two of their purposes, and only succeeds in their third because of a technicality in their wording.
I've never felt the need to join Mensa because I studied CS and AI, and all my friends from there are probably about as smart as I am, so I already have plenty of people to relate to, but she was always very interested in sport, which probably attracts a different crowd (though she's working in IT now).
Eventually I took the quick test which suggested I should take the longer test as a young adult looking to bolster my CV (résumé). Narrowly missed out and before I was able to submit again I had enough on there not to give it much thought again for me. An IQ test was part of the qualification process for the first job I took out of university, something I've not seen since.
I think my eldest son's probably at least as bright as I am, so this sort of thing has been on my mind again. We try and give him a breadth of experience - he does a lot of rugby (completely non-contact at his age), running and other sports, some child adventure centres, music lessons, scouts soon, etc. But I don't see too much wrong with adding Mensa as a small part in the context of a larger picture.
Yes, many practical sampling methods are biased, but it is not difficult for me to imagine that any result here is caused entirely by these biases.
but if you think the physical disabilities are better explained by something else to do with this population sample than their iq, you could very clearly publish this theory