Thursday, July 25, 2019

Four Saved Links

These were saved consecutively in September 2017 and all involve the same topic: IQ, including Richard Feynman's supposedly only above-average score. I just wrote about spatial intelligence, which is not formally tested as often or as well at school, and it may be that Feynman hit a ceiling on that but did not top out on other subtests, depressing his overall score a bit. That would also make my scores look better than they should. That makes sense to me, as I always felt they were a little higher than what my personal experience would suggest.  My numbers suggest I should almost never encounter people significantly smarter than myself, but I meet them all the time. I do look for them, but still.

You can get a deep dive on the subject of IQ here, especially if you follow the links.

Feynman, Schwinger, and Psychometrics. This is from Steve Hsu's Information Processing site, when he was still teaching physics at Oregon.  He is now at Michigan State as an administrator of (checks DuckDuckGo) of research and graduate studies.

Scott Alexander trying to get people to calm down. Slate Star Codex usually does first-class work. Quote: "So I want to clarify: IQ is very useful and powerful for research purposes. It’s not nearly as interesting for you personally." Amen and amen.

A Polymath Physicist on Richard Feynman's "Low" IQ and Finding Another Einstein. Steve Hsu again. Some of the same information, but mostly different.

Flynn Flips:  IQ Tests Do Matter. The "Flynn Effect" has been an uncomfortable array of data over the years for those who see genetics as the dominant driver of intelligence, as I do.  It does go against conclusions we might draw from other research, but it must be accounted for if we want to understand what is really happening.  Steve Sailer partly reconciled the two viewpoints over a decade ago. I probably should look around and see if there are any major changes over the last decade or so.  I do know that BGI (that's Steve Hsu again) has begun to uncover specific SNPs associated with increased IQ - about 50, last I heard - and as predicted, they each seem to have a tiny effect.  There is no collection of "you're really smart" genes out there.  Last I heard.


David Foster said...

Okay, if we take the farm laborer (in the Flynn Effect link) and compare him with, say, a literate mechanic in 1920, then clearly the mechanic gets more mental stimulation in a variety of ways.

But if we compare the 1920 mechanic with a 2019 fast food worker who is a high school graduate, the answer isn't so obvious. The mechanic has to do serious cause-and-effect thinking to diagnose and fix problems, and he also reads the newspaper, if only the sports and crime sections and the comics--and likely much more.

Whereas our current equivalent has no real need to think on the job comparable to that of the earlier mechanic; his job has been thoroughly Taylorized. And in many cases his K-12 education would not have given him much mental stimulation, either.

Assistant Village Idiot said...

I agree about it as better slow thinking, and that it is abstract and therefore more difficult. But the modern teenager has to think very fast as a fast-food worker. Not everyone can do it. I long-ago wrote about the phenomenon after visiting Europe, and noticing that the adults did not have the decision-making capability of American teenagers.

Urban environments seem to be more stimulating, and when correcting for race, those kids seem to do just a touch better. I really wish that were not true, and there may easily be a genetic factor or selection bias in those results.

BTW, the Romanians seem to have adjusted to making decisions for themselves, especially the young. I would still put American teenagers up against European teenagers in the workforce, however, except maybe the Scandinavians, whose work culture is intense.

This will come up again when I post my "most-visited posts" and get to the Education pieces.

RichardJohnson said...

A Polymath Physicist on Richard Feynman's "Low" IQ and Finding Another Einstein. Steve Hsu again. Some of the same information, but mostly different.

Which basically says that Feynman tested off the charts in math,but wasn't very good at verbal. I had a similar lab partner. He took Physical Chemistry and Organic Chemistry at the same time- VERY demanding courses, and aced them both. Not many can do that. But his writing skills were not good. I spent a lot of time correcting his lab report writing.

Texan99 said...

Once someone achieves great things, there's no longer much use talking about how to calculate his theoretical potential for achieving great things. If the calculated potential doesn't match the achievement, all it tells us is that our tests are flawed or incomplete, though they may be very useful statistically.