James R. Flynn recommends this paper, by Fox and Mitchum, in his new book:
Secular gains in intelligence test scores have perplexed researchers since they were documented by Flynn (1984, 1987). Gains are most pronounced on abstract, so-called culture-free tests, prompting Flynn (2007) to attribute them to problem solving skills availed by scientifically advanced cultures. We propose that recent-born individuals have adopted an approach to analogy that enables them to infer higher-level relations requiring roles that are not intrinsic to the objects that constitute initial representations of items. This proposal is translated into item-specific predictions about differences between cohorts in pass rates and item-response patterns on the Raven’s Matrices, a seemingly culture-free test that registers the largest Flynn effect. Consistent with predictions, archival data reveal that individuals born around 1940 are less able to map objects at higher levels of relational abstraction than individuals born around 1990. Polytomous Rasch models verify predicted violations of measurement invariance as raw scores are found to underestimate the number of analogical rules inferred by members of the earlier cohort relative to members of the later cohort who achieve the same overall score. The work provides a plausible cognitive account of the Flynn effect, furthers understanding of the cognition of matrix reasoning, and underscores the need to consider how test-takers select item responses.
The paper is here (pdf).
I credit video games, specifically, Doom.
+1
Agree – first person shooters require very good spacial reasoning skills. Today I can tell which of my friends played first person video games a lot based on their sense of direction – big game players can easily retrace their steps in an unfamiliar location (for instance), while others typically require a map.
IQ tests were out ahead of the rest of society in sensing which way the world was heading, which is why, remarkably, they’re still useful in their 2nd century.
“recent-born”: is that the American war on adverbs continuing?
Possibly.
DEATH TO ADVERBS.
I don’t understand a word of the abstract or of the first several pages of the paper. (I suspect that the problem is not entirely a consequence of any inability on my part to read and understand clearly written English prose.) Could someone explain to me what exactly their point is?
What remains to be seen is whether the Idiocracy Effect negates or merely reduces the Flynn Effect.
The Flynn Effect, named after Erol Flynn. Indeed another factor possibly helping advance IQs is better nutrician (sic) now than in the 1940s.
This has likely already happened. Brain case sizes have shrunk by about 10% in the past 20,000 years or so, and brain size is obviously correlated with IQ. However, if we were able to test throughout that 20,000 years, I would guess that IQ has gone up over time.
So the Flynn Effect may have occurred in spite of a mitigating idiocracy factor.
It varies by country. Height is another trait that had been increasing for a while, but now seems to be regressing in the tallest countries like the Netherlands.
As they switch away from nearly pure animal product based diet.
I’ve mentioned this before on here and elsewhere, but nutrition is surprisingly relatively unheralded as a candidate for the cause behind the Flynn Effect (I only saw it alluded to briefly in the linked paper upon skimming). It has been observed before in Norwegian conscripts that the Flynn Effect stopped alongside increases in height; height and IQ are both driven by nutrition.
Height and IQ have a lot in common. Height and IQ are correlated across and within populations. Both have a strong hereditary component (especially in western nations) and a strong component unexplained by genetics, whether it be randomness, environment or whatnot. Nutritional gains improve both, but have strong diminishing marginal returns. Height increases have stalled in western nations, as have IQ increases.
Great comment
Has nutrition improved at all over the last 40 years in say, the U.S.? If anything I would suspect it has gotten worse.
In the past 40 years the ‘obesity crisis’ occurred. Another term for obesity is over-nourished. We’ve also gotten taller over the last 40 years, which is almost definitely nutrition.
Nutrition is not calories, is it? I find it very hard to believe that an obese person has a higher average IQ than a slender person. Giving everyone 500 extra calories of sugar per day is not going to make them smarter.
@Cliff–just based on my own anecdotal evidence the smartest guy in class was always the fat kid, Tyler Cowen excepted.
Well we could ask John Pilley and Alliston Reid what dog chow they fed Chaser, the collie taught the names of 1022 objects, and suggest Arne Duncan incorporate it into the U.S. school lunch program.
Do you think Arne would follow through?
I’ll remember that next time I’m looking up at towering Jews and Asians.
Love this. Height and IQ correlate, huh? Explains all those brilliant basketball players. Lmao.
Love this. Don’t understand sampling bias or know how to use Google, huh? Lmao.
I’ve wondered whether the increased standardized testing of students would bump up results on IQ tests. If nothing else, familiarity with the format and test-taking strategies in general would evince some kind of advantage, I would think.
“nutrition is surprisingly relatively unheralded”: Mr Sailer frequently alludes to his belief that nutrition affects IQ.
I can’t say I fully understand the abstract, but the emphasis on analogies seems promising. And that brings up the question of: how wise was it for the SAT to get rid of its analogy questions about a decade ago. Not very, I would guess.
I mean a lot of what I do for a living is point out analogies that ought to be pretty obvious but few others have noticed.
The bottom line is that since the U.S. began rigorously selecting for its cognitive elites, it’s experienced a slowdown in economic growth. The industrial revolution didn’t need the SATs. Maybe it wouldn’t have happened with standardized tests and credentialism at all. Likewise, the Ivy Leagues were better off when they were the still havens of less than meritocratic WASPS.
IQ is obviously important, but once its abstracted out of its proper context, it becomes counter-productive. Centuries of high IQ Chinese wasted their cognitive gifts memorizing Confucian texts just like their Jewish counterparts squandered their minds studying the Talmud. Today, cognitive elites spend their most intellectually fecund years accumulating credentials and playing academic politics.
Ranjit – There are still plenty of high IQ people studying Engineering, Maths and Science, actually more than any other time in human history, so if there has been a slowdown in growth it is not due to all the high IQ people ending in rent seeking jobs (lawyers, investment banks, academics etc). If nothing else, there is a limited number of rent seeking jobs and high IQ people with low verbal skills will not want those type of jobs anyway.
Also, what slowdown in growth? The stagnation thesis is about middle incomes, not overall growth which hasn’t slowed down at all. The recent increase in inequality is probably explained better by improved meritocracy, which is opposed to your thesis.
Don’t they spend their most fecund years doing research and accumulating published papers so they can get tenure?
Excellent comment. See Peter Thief’s foundation that pays brilliant kids to stay out of college and develop their ideas.
On the other hand, Steve, much of your career is also pointing out how circumstances are NOT analgous (e.g. “These people are not, in fact, exactly like these other ones and those other ones are not a good model for them.”).
So people now are better at modelling things as some other thing, but not necessarily at choosing appropriate models, nor building new models. Sounds about right to me.
Within the realm of politics, for example, we more often conflate dissimilar things than did our predecessors. Attempting to achieving greater consistency (uniformity?) by ignoring differences is the modern way!
‘so-called culture-free tests’
Well, OK.
‘We propose that recent-born individuals have adopted an approach to analogy that enables them to infer higher-level relations requiring roles that are not intrinsic to the objects that constitute initial representations of items.’
And if this isn’t a definition of culture (learned behavior shared among many individuals), I’m not sure what is.
Score another point for those able to claim how something is ‘culture free’ with a straight face, unaware of just how culture bound they are.
Flynn’s book ‘What Is Intelligence?’ is very good and goes into these issues in some depth. I won’t attempt to sum it up in a blog comment.
cym plus roles that are not intrinsic to the objects that constitute initial representations of items.’
And if this isn’t a definition of culture (learned behavior shared among many individuals), I’m not sure what is
If the Flynn effect is really due to a learned skill, then there’s an easy prediction that can be confirmed simply by waiting: The growth in intelligence test scores will level off as penetration of the skill becomes saturated in the population. When everyone knows how to solve higher-order analogies, there will be no further increase in scores.
But what’s remarkable to someone like me who’s been away from this field for a long time is how archaic the whole approach seems. It’s incredibly superficial – what physicists would call “phenomenological”, or what economists would call macro without micro-foundations.
A 21st-century theory of general intelligence would be very different. It would start with the fact, known since Turing & Church in the 1930s, that there is a class of problem solvers that (“who”) are “universal” in a very strong and precise sense, and that any system (or any person) who can follow written directions and read and write their own notes about the problem in progress can accomplish any feat of pattern recognition that can be defined. It would then go on to observe that Kolmogorov-Chaitin complexity is relativized to a set of hardware axioms, and develop a theory of identification of hardware properties from observed relations between problem scaling and performance scaling, and then map those inferred hardware properties to brain properties. Cognitive psychology and cognitive science have already sketched out much of those mappings, with concepts like short term memory and procedural memory. Neuropsychology and related fields have identified what computer engineers would recognize as specialized graphics processors in the back of the brain, and digital signal processors in the temporal lobes, and general-purpose executive processors in the front, with most of the interior of the brain occupied by a switching fabric. We would expect that a complexity-based hardware identification theory would be able to rediscover these functional specializations based on data from tests such as Raven’s matrices alone.
Recognition of the higher-order relationships that Fox & Mitchum identify are just the tip of an iceberg of structure in problem solving capabilities, where more and more people have the implicit capability to solve problems involving such extreme abstractions as the Y combinator, and simply need to be taught how to realize their potential.
The most important I think it´s how the IQ is used. Doesn´t matter if you have the highest. By the way, the comment about height and his relation to the IQ is very interesting. I will read more about this. Have a great day. Kisses, paola.
Comments on this entry are closed.