Disaggregating the Flynn Effect?

Overall, the results support co-occurrence theories that predict simultaneous secular gains in specialized abilities and declines in g.

NB: this is for memory tests alone.  Here is the paper, via Rolf Degen.


Yeah. The Flynn effect represents environmental gains in IQ, but we know that fertility has been generally dysgenic in Western societies for the last 100 years, causing decline in genetic intelligence.

Western fertility may indeed have been generally dysgenic for the past century, but fertility rates themselves have fallen and our noble elites favor immigration over reproduction as a means of replacing the population.

But I'm sure that has had no effect on g either...

Your understanding of behavioral genetics is flawed.

Socioeconomic status modifies heritability of IQ.


That paper has zero implications on what's discussed here.

Flynn effect is caused by changes in the heritability of IQ due to environmental variation. Pretty simple to connect the dots.

No, that doesn't make any sense.

"is caused by changes in the heritability of IQ due to environmental variation."

What? I think you'd need a whole lot of evidence to back that claim up.

That Turkheimer paper is old. A meta-analysis of studies of the phenomenon it purports to find determined that the effect doesn't really replicate, especially outside of the US: http://journals.sagepub.com/doi/abs/10.1177/0956797615612727. Also, see the following: http://emilkirkegaard.dk/en/?p=6271.

> but we know that fertility has been generally dysgenic in Western societies...

No, we certainly don't "know" that. In fact the evidence seems to suggest zero to slightly positive correlation between IQ and fertility in the 20th century.


That is consistent with dysgenics overall and there is now genomic evidence for it, e.g. http://www.biorxiv.org/content/early/2017/06/06/146043.

>citing Jayman

Are you out of your mind? The guy's an anonymous blogger clown.

Genetic selection against g has been found is Iceland (http://www.pnas.org/content/114/5/E727) and the US (http://www.pnas.org/content/113/28/7774.abstract). This isn't up for debate anymore.

Maz is known eugenics poseur PumpkinPerson.

I am not but PumpkinPerson does appear to be closeted.

"Overall, the results support co-occurrence theories that predict simultaneous secular gains in specialized abilities and declines in g."

What? Is that saying that learning stuff reduces g?

No. The heritable components of intelligence (g probably being most heritable) seem to be decreasing due to dysgenic fertility. At the same time, improving environmental conditions have allowed negligibly heritable components of intelligence to increase.

This paper supports what the hereditarians have been saying when the Flynn effect is brought up in the context of (and as an attempt to refute the importance of) racial iq differences.

I don't get that. "Flynn effect" is just used here as increase over time (or anti-Flynn decrease over time). They are not working with IQ. This was a look at how other mental facilities fare over time.

And certainly no discussion of"race" or trajectories of improvement "by race."

What good is g if it isn't giving you a pretty broad picture of intelligence, including so-called "specialized abilities"? Isn't that the whole point of g? Maybe this is a reason to think g isn't all it's cracked up to be?

My guess has been that the Flynn Effect has been related in part to Moore's Law, or to something as of yet unnamed of which Moore's law is a subset.


I could see a measurable difference when we send the best and the brightest off to wars that incur high mortality rates. The genetic reserve would exist with the women and some men left behind. Now that women are in the military, if a conflict occurs that requires greater military staffing due to higher mortality rates, that genetic reserve may be at risk as well.

Seen a lot of discussion of this paper but almost no one, except the authors themselves, is crediting Michael Woodley for having originally proposed the idea that dysgenic decline on g, called the Woodley effect by Charles Murray, is co-occurring with gains on specialized abilities, aka the Flynn effect. The co-occurrence theory was put forward in a 2012 paper of his, "A life-history model of the Lynn-Flynn effect" http://www.sciencedirect.com/science/article/pii/S0191886911001498. He also found evidence of the Flynn effect and dysgenics in short-term and working memory tests in 2015 http://www.sciencedirect.com/science/article/pii/S0160289615000410. The authors of the paper Cowen posted just replicated his findings in a much larger data set and with Corsi block as well as digit span memory tests.

This has been known for decades and is [obviously] not due to Michael Woodley, who if memory serves also had wildly ridiculous ideas about Victorian intelligence.
Perhaps it should be noted that [obviously] the Flynn effect was not really originally discovered by Flynn. Anybody who was a serious psychometrician knew that there were increases on [certain subtests of ] IQ examinations.

You have no idea what you're talking about. The authors only cite Woodley when discussing the co-occurrence theory because he came up with the idea. The co-occurrence theory does not simply posit that there are only gains on certain subtests of IQ tests. Rather it maintains that different components of intelligence have increased and declined over time. g has gone down while specialized abilities have gone up. Prior to Woodley's work, psychometricians, like Richard Lynn and James Flynn for example, argued that improvements of environment simply overwhelmed dysgenic fertility at the phenotypic level, leading to an overall increase in IQ--that was the end of the story. They did not argue that g itself was declining even as overall IQ test scores were rising. That was Woodley's idea. You are correct that Flynn was not the first to discover the Flynn effect, but are apparently too dense to understand that the co-occurrence theory and the Flynn effect aren't the same thing.

As far as Woodley having "ridiculous" ideas about Victorian intelligence goes: well, subsequent analyses have shown that he was right about reaction times, even after controlling for various factors noted by critics (http://journal.frontiersin.org/article/10.3389/fnhum.2015.00452/full) and even in modern cohorts tested with exactly the same methods over time (http://journal.frontiersin.org/article/10.3389/fnhum.2016.00407/full). Also, a study of declining IQ-related polygenic scores over about a century in Iceland indicates an IQ decline of about 1 point per decade, as Woodley predicted, once a reasonable estimate for the heritability of IQ is used: http://www.pnas.org/content/114/5/E727. Note there are other factors lowering IQ, other than genetic selection, that is, which is all the PNAS paper examines, the effects of which would be reflected in changing reaction times. So the actual IQ decline will be even bigger than the Kong paper suggests, even after a more reasonable IQ-heritability estimate is applied, once the other factors are taken into account.

Comments for this post are closed