a better way to raise average IQ is to encourage smart people to breed?
That's an interesting hypothesis, but the effect size of environmental changes is MUCH larger than the effect size of changing gene frequencies in increasing the IQ scores of the general population. It has been typical for decades for authors who take a strongly hereditarian (which means "pre-Mendelian," really) view of influences on IQ to suppose that IQ score trends over time would be for the population mean to decline. In fact, population trends over time in countries all over the world have been HUGE increases in IQ.
The Stanford University researchers who developed the first few versions of the Stanford-Binet IQ tests had access to data for decades that could have revealed a surprising trend in raw item content performance on IQ tests. The trend, now known as the Flynn Effect, after Professor James R. Flynn, who had the greatest role in discovering it, is that raw scores on IQ tests, including all kinds of IQ tests loaded on fluid intelligence, have risen over time in national populations all over the world.
James R. Flynn started slowly but worked steadily in finding data sets and eventually found a large body of tests for which raw score data is available over long time series. His first published paper (1984) on the issue of IQ score changes over time was based on the renorming of the most commonly used IQ tests in the United States. He found that as new editions of the major IQ tests (the Wechsler test and the Stanford-Binet) were published, that the newer edition of each test invariably had a norming sample population that did better on the old edition's item content than had the old edition's original norming sample population. In other words, for a given brand of IQ test, as time went on, the average person did better and better on the raw item content of that test, resulting in higher and higher IQ scores over time for people tested on the same test with score calculation based on the earliest norms.
The possibility that raw scores on IQ tests were rising so consistently, and in such magnitude, was so surprising that it prompted many psychologists to cast doubt on the sample populations used for the early studies that showed this phenomenon--even though those sample populations were none other than the norming sample populations for the major brands of IQ tests. Psychologist Arthur Jensen proposed to James R. Flynn in January 1983, while Flynn's first major article on score trends was awaiting publication, that a good data set to show changes in IQ scores over time should
a) be comprehensive, e.g., a test of an entire national population, to eliminate the possibility of sample bias,
b) use the same test from generation to generation, with time trends shown by raw score differences,
c) emphasize "culture fair" tests such as Raven Progressive Matrices, which were presumed then not to include item content that is taught by compulsory schooling,
d) be based on adult populations, who have reached a mature level of intellectual functioning, to minimize the effect of differing rates of intellectual growth in childhood from one generation to the next.
Amazingly, Flynn found data sets with all of those characteristics, especially data sets from compulsory IQ testing of NATO draftees in the Netherlands, Belgium, and Norway. The data based on a subset of items from the Raven Progressive Matrices test showed substantial raw score rises for Dutch draftees: if the 1952 population mean is taken to be IQ 100, by 1982 the Dutch male mean IQ was 121. The IQ raw score rise over time was smooth, and by the end of this period had resulted in ceiling effects on the test for a significant number of draftees (Flynn 1987). Flynn has suggested an experimental design that might help unravel the causes of the increase in raw IQ scores over time, while reviewing newly discovered data sets that extend observation of IQ score rises backward in time to the beginning of IQ testing. The Raven test, early in its use, was given to an age cohort born in the 1870s as part of a study of IQ changes over the course of adult aging. Careful analysis of this cohort allows extending IQ raw score trends back to the beginning of IQ testing, showing that fully 90 percent of Britons born in 1877 scored below the fifth percentile of Britons born in 1967, or in other words that most turn-of-the-last century Britons had an IQ of 75 or lower on a highly g-loaded IQ test, according to current norms (Flynn 1999; Flynn 2000b).
As Mackintosh (1998, p. 104) writes about the data Flynn found: "the data are surprising, demolish some long-cherished beliefs, and raise a number of other interesting issues along the way."
CITATIONS:
Flynn, James R. (1984). The Mean IQ of Americans: Massive Gains 1932 to 1978. Psychological Bulletin vol. 95, pages 29-51.
Flynn, James R. (1987). Massive IQ Gains in 14 Nations: What IQ Tests Really Measure. Psychological Bulletin, vol. 101, no. 2, pages 171-191.
Flynn, James R. (1998). IQ Gains over Time: Toward Finding the Causes. In Neisser, Ulric (Ed.). The Rising Curve: Long-Term Gains in IQ and Related Measures. Washington, DC: American Psychological Association.
Flynn, James R. (1999). Searching for Justice: The Discovery of IQ Gains over Time. American Psychologist, vol. 54, No. 1, pages 5-20.
Flynn, James R. (2000a). IQ Gains, WISC Subtests and Fluid g: g Theory and the Relevance of Spearman's Hypothesis to Race. In Gregory Bock, Jamie Goode & Kate Webb (Eds.), The Nature of Intelligence (Novartis Foundation Symposium 233) (pp. 202-227). Chichester, United Kingdom: Wiley.
Flynn, James R. (2000b). IQ Trends over Time: Intelligence, Race, and Meritocracy. In Kenneth Arrow, Samuel Bowles & Steven Durlauf (Eds.). Meritocracy and Economic Inequality (pp. 35-60). Princeton: Princeton University Press.
Mackintosh, N. J. (1998). IQ and Human Intelligence. Oxford: Oxford University Press.
If the Flynn effect was true, then our great-grand parents must have had average IQs that would have categorized them as mentally disabled today. Clearly, this was not the case. My grand parents in their prime were just as smart as my parents, and my parents and uncles are just as smart as my cousins.
It's also interesting that measures of vocabulary and numeracy have been flat over the past century. More evidence that cognitive abilities have been flat.
emphasize "culture fair" tests such as Raven Progressive Matrices, which were presumed then not to include item content that is taught by compulsory schooling,
Here is the problem. Doing a Raven Progressive Matrix is just as much a learned skill as math or vocabulary. If people went from learning no math in 1900, to learning lots of math in 2000, scores on math tests would go up, even without any gains in underlying cognitive abilities.
All the Flynn effect means is that people have much more exposure to the type of puzzles in the Raven Progressive Matrix today than they did in the 1950's. Numeracy and vocab are actually better measures of historical trends in IQ, because environmental exposure has been more constant between the two time periods.
When people try and measure intelligence, they are trying to measure the underlying cognitive abilities that allow people to learn faster and achieve higher levels of cognitive skills. Unfortunately, there is simply no way to measure this directly. The only way to measure it, is to give a person a skill test, and then control for environment. This will always be fraught with potential for error.
My grand parents in their prime were just as smart as my parents, and my parents and uncles are just as smart as my cousins.
And this is one of the best empirical proofs that IQ scores are not a "measure" of how smart someone is. James Flynn discusses exactly this point of yours in several of his writings. He relates a story from Arthur Jensen about a mentally retarded man who claimed to be baseball fan, but who was very vague about the rules of baseball and didn't seem to know the names of many professional players. Yet that man had an IQ score that would relate back in time to a population average score from the era when baseball became a popular sport, widely followed in the United States.
All the Flynn effect means is that people have much more exposure to the type of puzzles in the Raven Progressive Matrix today than they did in the 1950's.
I'm very sympathetic to this statement, because I used to think that it offered the best explanation for the Flynn effect. But I am now convinced by Flynn's latest book
that on the one hand the gains in IQ test scores are real, and not just artifacts of familiarity with test item content (in large part because so many different kinds of tests have all shown this effect) and on the other hand that IQ has increased in society, and has been applied in the labor market and other aspects of daily life, without wisdom (Flynn's term) or rationality (Stanovich's) term increasing as generally in society.
I haven't read the books you've mentioned, although I'm pretty familiar with the debate, and the latest arguments.
I think it's interesting to make an analogy with a sport like running. If you measure running ability in a group where every person has absolutely no training, all the differential in running ability will be genetic. But a person will be able to train and improve, and so a change in environment ran result in gains in ability. If some people in the group start to train harder, then the differential in scores will be a mix of environmental and genetic.
If you compare members of a cross country team who all endure the same grueling training, the differences in their running speeds will be genetic. If one member tries to train even harder, he will probably be unable to raise his speed even higher. At some point, he maxes out and no training will raise his speed even further.
The subtests that have a flat trend were vocab, math, and general knowledge. What this indicates is that the environment of a century ago maxed people out at these skills a century ago. This is not surprising, as these skills are the most culturally common skills. Even in 1900 a person had constant exposure to reading, math, and general knowledge.
The subtests that have seen the greatest rise are all obscure skills - object assembly, picture arrangement, similarities. My guess is that the great increase in schooling resulted in more training in these obscure skills. Increased availability of games and puzzle books also helped. In 1900 if you were a genius kid, you read the classics. In 2009 you can play all sorts of fun puzzle games online or in books. As a result of this, scores on these obscure puzzles have gone up.
The net of this may mean that the hardcore hereditarians ( Murray, Jensen) are wrong about the specifics but right about the big picture. They are wrong to think that differences in the Raven test by themselves mean that one group has more natural cognitive abilities than another group. But they are probably right that individual and group differences in the economically and socially useful cognitive skills are mostly genetic. The differences in the obscure cognitive skills ( Raven's, picture arrangement) may be much more environmental.
I suspect that the Flynn effect begins and ends with the improvement to diet, as is the case with height. Though really, society has changed in so many ways in 100 years, that it's impossible to say this or that was responsible, therefore let's replicate it. At any rate, there's a serious upper limit to it.
I suspect that the Flynn effect begins and ends with the improvement to diet
Flynn points to data from a World War II famine in the occupied Netherlands showing that the famine region had no difference in IQ changes over time from the region that didn't experience famine. The IQ score increase has also occurred in countries that were well nourished to begin with for the most part.
I'll accept "leveled off" regarding certain countries, having read Flynn's work, and I'll accept "starting to reverse" if you are only talking about crystallized intelligence, for the same reason, but do you have any current citations to a disfavorable trend anywhere in fluid intelligence?
I'm looking for a copy of the paper that doesn't require access to a university archive. Every link so far goes to the same source.
the name of the paper if you want to look for it is
A long-term rise and recent decline in intelligence test performance: The Flynn Effect in revers
After all, this makes sense. Flynn effect seems to be working mostly by reducing "bad stuff" going on that would make you dumber than your potential (most of the gains are concentrated in the lower half of the distribution and negligible in the top half).
Once everybody's environment is more or less ok, there is no magic that will start to produce geniuses from normal people.
There was a comment on iSteve, where someone said that perhaps whites were not intrinsically smarter than the global average (of about 90) but, that they've had a headstart of a century of Flynn Effect.
this is an actual possibility, watching the flynn effect in devloping vs industrialized nations is what led to the idea (AFAIK)
there's not enough data to be able to draw any conclusions, partly due to the stigma of intelligence research.
We had almost this exact same discussion a month ago. Firstly, you quote only from Flynn and not from any scientist holding an opposing viewpoint (e.g. Arthur Jensen). Although Flynn is a good scientist, I don't think that it is good to rely just on the studies of one political scientists and activist – he is author of “How to Defend Humane Ideals: Substitutes for Objectivity” after all).
There are numerous studies that also show that the general intelligence factor is due to physiological effects (e.g. size of different areas of the brain, etc...). How do you explain that?
There is a theory on why when people are trained in a task they do better – it turns out that the amount of “general intelligence” used is less. There are also studies that show this effect with PET scans of the brain:
> Thus, the largest GMR decreases with practice were found on subtests with the highest
g loadings. Jensen noted this finding was consistent with what he termed the conservation of g.
Namely, with practice and training, tasks become more automatized and require less g.
(from Haier [1])
There is a recent study (2008, te Nijenhuis ,van Vianen , van der Flier [2]) that shows that the increase in RSPM and other scores is because of training and is not an increase in the overall general intelligence [2] (the studies that you cited is all 8 years or older). Incidentally this study cites two of the Flynn papers that you cited (Massive IQ Gains in 14 Nations: What IQ Tests Really Measure, IQ Gains, WISC Subtests and Fluid g: g Theory and the Relevance of Spearman's Hypothesis to Race).
The idea that Flynn proposes (i.e. that general intelligence is increased over time is dismissed). Here is a full quote of the abstract of the paper (if you need the full paper and do not have access to it, msg. me and I will email it for you):
IQ scores provide the best general predictor of success in education, job training, and work. However, there are many ways in which IQ scores can be increased, for instance by means of retesting or participation in learning potential training programs. What is the nature of these score gains? Jensen [Jensen, A.R. (1998a). The g factor: The science of mental ability. London: Praeger] argued that the effects of cognitive interventions on abilities can be explained in terms of Carroll's three-stratum hierarchical factor model. We tested his hypothesis using test–retest data from various Dutch, British, and American IQ test batteries combined into a meta-analysis and learning potential data from South Africa using Raven's Progressive Matrices. The meta-analysis of 64 test–retest studies using IQ batteries (total N = 26,990) yielded a correlation between g loadings and score gains of − 1.00, meaning there is no g saturation in score gains. The learning potential study showed that: (1) the correlation between score gains and the g loadedness of item scores is − .39, (2) the g loadedness of item scores decreases after a mediated intervention training, and (3) low-g participants increased their scores more than high-g participants. So, our results support Jensen's hypothesis. The generalizability of test scores resides predominantly in the g component, while the test-specific ability component and the narrow ability component are virtually non-generalizable. As the score gains are not related to g, the generalizable g component decreases and, as it is not unlikely that the training itself is not g-loaded, it is easy to understand why the score gains did not generalize to scores on other cognitive tests and to g-loaded external criteria.
So what we have for the immutability of general intelligence is a theory, intelligence test results confirming the theory and PET scans of the brain showing this effect.
[1] Richard J. Haier, Ph.D., Positron Emission Tomography Studies of Intelligence: From Psychometrics to Neurobiology
[2] Jan te Nijenhuis , Annelies E.M. van Vianen , Henk van der Flier, Score gains on g-loaded tests: No g
PS: This reply is a little long so sorry for that.
I've read more of what Haier has to say since his name last came up in an HN thread. I especially like this quote:
"Just because intelligence is strongly genetic, that doesn't mean it cannot be improved. 'It's just the opposite,' says Richard Haier, of the University of California, Irvine, who works with Thompson. 'If it's genetic, it's biochemical, and we have all kinds of ways of influencing biochemistry.'"
The all kinds of ways to influence biochemistry include influences that are not themselves biochemical. Things that people DO can cause biochemical changes in their brains. (Sleep deprivation is well known to do that, for example, as are activities that bring about good mood.) Very likely engaging in challenging education brings about biochemical changes that result in changes in IQ.
you quote only from Flynn and not from any scientist holding an opposing viewpoint (e.g. Arthur Jensen)
First of all, let me say very forthrightly that maybe mainstream journalism operates on an equal-time principle, but science manifestly does not. It actually makes no difference at all who says anything in science, because it should be examination of the data that leads to a conclusion, not examination of a scientist's reputation. But since you asked me to quote Arthur Jensen, I will be glad to:
"Now and then I am asked . . . who, in my opinion, are the most respectable critics of my position on the race-IQ issue? The name James R. Flynn is by far the first that comes to mind." Modgil, Sohan & Modgil, Celia (Eds.) (1987) Arthur Jensen: Concensus and Controversy New York: Falmer.
> 'It's just the opposite,' says Richard Haier, of the University of California, Irvine, who works with Thompson. 'If it's genetic, it's biochemical, and we have all kinds of ways of influencing biochemistry.'"
There is good evidence that several compounds can increase intelligence (albeit temporary in most cases). A good example is several stimulants. The best way would probably be to alter the genetics directly. This however does not take away the fact that it is still innate and most people are not likely to see their intelligence changed.
> Very likely engaging in challenging education brings about biochemical changes that result in changes in IQ.
Is there any evidence that this is the case (i.e. research studies)? This may happen – but I have not seen any documented evidence of that this is the case.
If this was the case, children adopted at birth in middle class families would have general intelligence factors matching their step-parents (this is not the case). Also, the general intelligence factor would be changed with learning – but there has not been any evidence that this is the case.
> First of all, let me say very forthrightly that maybe mainstream journalism operates on an equal-time principle, but science manifestly does not.
That is true. Views of science with which the majority of scientists agree is generally represented as the opposite in the public press. A good reason for that is in the public sphere the “ought to” become “is” and scientists reinforcing that is given publicity.
There are several books like that (a lot of them by Gould).
> It actually makes no difference at all who says anything in science, because it should be examination of the data that leads to a conclusion, not examination of a scientist's reputation.
This is not entirely true. I am the last person to say a dissenting scientists should be silenced but the public simply cannot go into detail in any of several subjects. A good example is client change – no single person in the public is smart enough to know all facets of climate change science.
The public view is largely made up of the consensus view of a lot of scientists – but that does not mean that there is no place for the Bjorn Lomburgs.
The intelligence debate is the same – we should not rely on a single scientist. No single scientist is a master of all the fields (from neurobiology to psychometrics).
It is unfortunate that the politically correct view (the “ought to”) is promoted over the actual science. Public press is not representative in what goes on in the research.
> “Now and then I am asked . . . who, in my opinion, are the most respectable critics of my position on the race-IQ issue? The name James R. Flynn is by far the first that comes to mind." Modgil, Sohan & Modgil,
That is true – and Jensen is correct in that. Flynn is in many cases the most prominent proponent of many views. My problem with Flynn is twofold.
The first problem is that he is a political scientists and he tries to defend the “ought to” - instead of trying to search for the “is”. If he was just active in intelligence research (instead of politics) it would be fine.
The second problem is that several scientists near the intelligence debate is viscously attacked for where they get their funding (in a politically motivated attempt to cut of funding). Their scientific objectivity was questioned and it is really ugly. Flynn is active in politics – yet no one attacks his scientific objectivity. It is a double standard that is simply not right.
no one attacks his [Flynn's] scientific objectivity
Yes, because he has demonstrated his scientific objectivity by changing his point of view from time to time, digging up new evidence when scientists say his previously offered evidence is inadequate, and scrupulously honoring his most ardent opponents with credit when their counterarguments prompt him to reconsider his previous publications. What Flynn writes in the early twenty-first century about IQ is much better quality research than what he wrote in the 1970s. (It's important to point out that already by the 1980s he was being published in Psychological Bulletin, the most prestigious journal in psychology, because his articles were meeting a high standard of scholarship.) Take a look at which psychologists and sociologists praise Flynn, his research in general, or his latest book on the Amazon.com page for his latest book:
Ian Deary, Edinburgh University
Charles Murray, American Enterprise Institute & co-author of The Bell Curve
> Yes, because he has demonstrated his scientific objectivity by changing his point of view from time to time, digging up new evidence when scientists say his previously offered evidence is inadequate, and scrupulously honoring his most ardent opponents with credit when their counterarguments prompt him to reconsider his previous publications.
This is true for Jensen and several other researchers - yet they are treated as pariahs in the public arena.
Several have questioned the objectivity of Jensen (and other researchers). As I have said before, James Flynn is both a passionate left-wing politician and he freely admits that he is egalitarian. E.g. in “How to defend Humane Ideals: Substitutes for objectivity”:
> This book was written by someone committed to humane-egalitarian ideals...
So, how can anyone question Jensen's objectivity but not that of Flynn?
There are several attacks on Jensen and others holding up Flynn as the poster boy (see at the end). Flynn has not done anything to dispel that. He even wrote a book attacking Jensen indirectly (Race, IQ and Jensen). Here is a quote:
> However, the last stand of the racist is not without importance, something I will attempt to demonstrate by giving a racist ideologue his say.
Scientists should search for truth – with not prior preconceptions or beliefs (i.e. he should search for truth and not just justification).
Here is an example of some Jensen attacks:
“Resurrecting Racism: The modern attack on black people using phony science” (by the popular anthropologist Francisco Gil-White):
I will not quote it since the whole chapter is full of Jensen attacks.
The extremely popular “The Mismeasure of Man” by Stephen Jay Gould also have the same attacks Jensen and others.
The list goes on. There is a saying that I translated to English: “They are not sheared over the same comb” i.e. Flynn are treated differently than those with opposing viewpoints.
It actually makes no difference at all who says anything in science
We aren't talking about physics where the variables are defined precisely. We are talking about psychology which is a pseudo-science at best as there are so many components that one may not really know what caused what or if it was something else that caused it. So when it comes to pseudo-science, a place where research is mosaic with many studies supporting the hypothesis and many others not, there is plenty of room for authors bias.
As for the debate of nature v nurture you to are having, one of you summed it up pretty well, the environment causes changes in the brain, but the question is a bit circular, namely did the brain make the person seek out the environment which led to the changes of the brain, or was the environment unrelated to the agents intentions. Anyhow, the question is not which one it is, but to what extend does nurture or nature influence. So its proportionality not absolutism.
