Get Smart

A while ago, I read this Nicholas Kristof column regarding the intersection of poverty and intelligence in the NYT. I thought that it was interesting, but a little off, and didn't think more about it. So, Westy forwards me the link, probably based on some of our previous discussions on the nature of expertise, and the impact of race/poverty on higher education. I started to write an e-mail, that I realized was running long, so I decided to post some thoughts here.

A Broader Thought
A reasonable explanation for intelligence is the interaction of natural ability (genetics) and one’s environment. Generally, that’s a reasonable explanation for just about everything. Think Elijah Dukes, a volatile baseball player who we've been talking about over in the UPL - lots of ability (power, speed), probably some negatives with genetics (dad was a murderer), and likely a lousy environment (dad was a murder). I haven’t read the Nisbett book (though I’m familiar with his research), but the Kristof piece paints too rosy a picture. There’s a limit to potential growth (i.e., environmental benefits to IQ), and when you sell the idea that there isn’t, you create all sorts of issues, particularly over-inflating self-esteem to the point of entitlement. Interestingly (and probably tangentially) I suspect that there’s a bias toward socialist solutions, rather than solutions that push individuals to do more. His writing:

“Good schooling correlates particularly closely to higher I.Q.’s. One indication of the importance of school is that children’s I.Q.’s drop or stagnate over the summer months when they are on vacation (particularly for kids whose parents don’t inflict books or summer programs on them).”

His statements suggest that no matter how good the schooling is that there’s fadeout (the IQ drop). It’s parenting (pushing individuals to do more) that keeps IQ up (there was a Freakonomics posting on this, at some point, right?). But the next thought pushes these intensive schools (changing the system). I think that’s a bad leap of logic. My interpretation of that fact is that we need to educate the parents. And if I were writing an article on the issue, my next sentence would be about how there should be more of an emphasis on the interaction of parenting and education, not the need for more intensive schools. But I suppose that's a matter of philosophy. If you believe that the system needs to be changed at a more macro level, then you change schools. If you believe that individuals need to be pushed in their micro environments, then you push interventions that breed parental involvement.

More Specific Thoughts
One inherent issue is what IQ actually measures. Generally, IQ measures intelligence, which is the way that Kristof has written up the article. However, I think that most folks who study intelligence break intelligence down into at least 2 dimensions (the dichotomy that I have described in the past is this fluid vs. crystallized intelligence). I think that it’s more accurate to say that genetics correspond to fluid intelligence, whereas the environment heavily drives crystallized intelligence. So there's one major issue there. But where are the changes in IQ coming from? Again, I’d like to see the data on that.

Moreover, the use of statement's like, “Another indication of malleability is that I.Q. has risen sharply over time. Indeed, the average I.Q. of a person in 1917 would amount to only 73 on today’s I.Q. test. Half the population of 1917 would be considered mentally retarded by today’s measurements,” drive me nuts. It's, at best, naïve, and at worst, disingenuous. I.Q. is a measurement of a construct that's based on intelligence/ability. I.Q. is defined such that average I.Q. is 100, and the standard deviation is 15. So, if you were to take the entire population in 1917, and have them take any I.Q., test, the average would be 100, and the standard deviation would be 15. I.Q. must stay the same, by definition. Now, I think that the intended idea is that the average amount of knowledge and analytical capability has increased since 1917.

Does this make sense? If half the people in 1917 were retards, then probably another half, or about 3/4 of the population, 90 years before that (in 1827) were also retards. And about 7/8 of the people in 1737 were retards. And about 15/16 of the people were retards in 1647, etc... how in the world did we ever evolve? We should be sitting around a pile of unlit wood, dragging square boxes along the ground, and trying to club elephants for food. How do you explain the wisdom of classic literature and religious texts?

I'd quibble a little about the amount of knowledge that differs. The amount of knowledge that people have today should be more than what people had 90 years ago. There has been a great advancement in the amount of knowledge that is known by man. So, it follows that this should trickle down to what each individual knows. However, I'd argue that how much an individual actually knows doesn't really change a whole lot. The human brain can do only so much. Things like short-term and long-term memory are generally similar, across individuals (at least in the same order of magnitude). I'd say that what is important to know has changed quite a bit. If you put that corresponding expectation of knowledge on a test, then of course, people of different eras will do poorly. But that can generally be learned pretty quickly.

And more problematic is the raw cognitive ability that people possess. Things like literacy and numeracy were not as prevalent 90 years ago. But does that account for raw cognitive ability? Or are IQ tests a function of the ability to navigate the written world? And do we really think that is the only form of intelligence? I don't know, but when I look at this sort of article, I just roll my eyes. How do you measure the intelligence required to domesticate animals or to develop agriculture? We don't do that particularly well.

Personally, when I read about intelligence and human ability, I prefer things like this classic article on intelligence.

-Chairman