Notes On the Flynn Effect

[nyuanshin]

The Flynn Effect has to be the weirdest finding in psychometrics. If you take it at face value, it suggests that someone in the bottom quintile of intelligence today would be in the top quintile of intelligence at the turn of the 20th century. I have no trouble believing that people were dumber on average back then, but that much dumber? It just

Comments

[smandal]

The Wicherts et al. themselves mention the possibility (last paragraph of conclusion) that intelligence might not be characterizable by latent variables. This implies that at the end of the day g is all that matters, manifesting quite diversely in different cohorts.

Also, height is generally a dominant trait, so those having this trait would benefit more from nutritional gains than those who don't, explaining why the upper side of the bell curve has moved instead of the lower. (Anecdotally, this is quite apparent in the children of Indian immigrants to North America.)

[nyuanshin]

Good point about height.

Either I'm misunderstanding you or you misunderstand the Witcherts paper's conclusion. The g factor *is* a latent variable, and they don't suggest that intelligence can't be characterized by latent variables (that would a lot more radical than what they're really saying); they're saying merely that the *gains* in IQ can't be explained solely by changes in latent variables.

[smandal]

Perhaps I read too much into this statement (penultimate paragraph): "Differential gains resulting in measurement bias, for example, imply that an overall test score (i.e., IQ) changes in composition."

So, while the first principal component may rise only slightly, the IQ score given by a particular test might rise drastically -- this is precisely what you wrote.

[nyuanshin]

Ahh. Yeah, that sounds right to me.

[cobalt999]

I'm onboard with the correlation to nutrition, and agree the link seems fairly convincing, but I wouldn't dismiss the effect of widespread public schooling. The Flynn effect could very easily be the result of both, given how our neural pathways are shunted and shaped from birth through the end of adolescence. Nutrition facilitates the multiplication of neurons in early development, and perhaps schooling acts to reinforce pathways friendly to IQ tests. The developing cerebrum is pretty dynamic; without early demand for abstract reasoning, corresponding pathways may atrophy. The threshold for meaningful effect may be low, by our educational standards, and achieved early

[nyuanshin]

The problem with this is that you have to posit qualitative improvements in early schooling over the past 80 years that dwarfs the variance in early schooling quality within currently existing schools. I'm not sure how plausible that is. In any case, this has been tested: With kids in early-intervention programs like Head Start, you see a brief advantage in intelligence test scores which then fades into nonexistence by the time they hit 4th grade. Also, in the Kenya study both cohorts attended the same schools with much the same conditions.

If we're looking at early influences I'd look to declining family size (something I forgot to mention in this post) before I'd look to schooling: When parents can spend more time with each individual child the child is likely to get more stimulation very early on in life. This would also cohere with findings that firstborn children tend to have slightly higher average IQs than later-born children.

[cobalt999]

From my reading, the Kenya study seems to point out a significant increase in preschool education between the first and second cohorts; this corresponds to the early developmental period I'm talking about. In fact, I'm kind of surprised they didn't compare the cognitive measures of the siblings who attended preschool to those who didn't, within each cohort

[nyuanshin]

Fair points about the Kenya study; we need more and better studies from other developing nations explicitly designed to tease out the relative influences of nutrition and early stimulation.

Bear in mind that nutrition isn't just about raw calories-micronutrients like iron, iodine and folate matter a lot for brain development in the womb and early development. A study in Spain found that presence of these nutrients corrlates modestly with IQ even within the normal range of a developed country. Also bear in mind that there's an economy at work in your metabolism: a lot of different organs require the same stuff, and if there's a tradeoff between (say) a more robust kidney or heart or intestine and a slightly bigger brain, your body will usually go with the essentials and sacrifice the luxuries. You can't synthesize extra fatty acids or micronutrients that simply aren't there.

Here's a pretty good meta-analysis of early-intervention programs finding that they do confer benefits on academic performance but not on IQ

[lestp]

i think it's not (only) the general nutrition that is responsible, but refined sugar and other glucose-rich foods
we have more and more of it in our diet
and this stuff is pure brain fuel
no processing, no nothing, goes straight to the head
if you have too much of it, you get sugar rush
but if you get almost too much but not quite, you can have your brain running in overdrive around the clock
so, the brain is the same as 200 yrs ago, but it's running on high-octane fuel instead of kerosene

Music and the "school of life" hypothesis

[soulchanger]

One way in which society has changed to favor complex mental stimulation, even in the very young (in fact, even in the womb), is the increasing availability of recorded and broadcast media. We don't usually think about the complete saturation of our daily lives with music - with pop music, commercial jingles, showtunes, etc - because we were all born with it, and have been exposed to it since birth

Re: Music and the "school of butts" hypothesis

[nyuanshin]

Yeah, Flynn has alluded to these media effects as a possible contributor and Steve Johnson has popularized the idea. I don't discount it completely, but the evidence for it is really lacking-the "Mozart effect" turned out to be mostly horsepuckey, for instance, and most of the evidence for critical periods in development has been overblown.