[THS] Is America smart enough? IQ and national productivity.

[Peter Webster]

Is America smart enough? IQ and national productivity. (Cover Story)
From: National Review  |  Date: 4/15/1991  |  Author: Seligman, 
Daniel 
 
Arguments about group differences in IQ have roiled the academy for years. Oddly enough, nobody pays attention to certain national IQ differences, which could help us answer the question: How does Japan do it?

INTELLIGENCE testing has a bad, bad reputation these days. At least, it does among the educated classes. Having spent the past three years working on a book about IQ testing, and having chatted extensively with assorted friends and publishers about the project, I vouchsafe that American college graduates overwhelmingly believe all of the following: IQ tests are culturally biased. Test-score differentials reflect mainly the advantages of the affluent. The tests discriminate in particular against minority-group members. They unfairly stigmatize low-scoring children. And in any case IQ tests do not measure anything it might be reasonable to label "intelligence."

This highly disparaging view of IQ tests tends to be asserted confidently and sustained with considerable passion. It causes book publishers to run for the hills when confronted with a "pro-IQ" manuscript (as I have sorrowfully discovered). And it continues to prevail in public discussion even though most experts-the psychologists and other scholars most closely involved with IQ test data-hold every one of the propositions above to be false.

IQ test data have enormous explanatory power. Far more than most socioeconomic variables, the data explain why some people get ahead and others do not. Indeed, their explanatory power is the main reason for their broad unpopularity. Egalitarian America wants its children told, You can be anything you want to be." The message from the testers is that this is pious baloney. This message is especially unwelcome because it disproportionately affects blacks and Hispanics, whose IQs on average are significantly lower than those of whites.

The IQ controversy nowadays is heavily centered on the meaning of these group differences. This article represents an effort to move the controversy into some less familiar territory-to extend the discussion to national IQ differences, and thereby explain (well, partly) why some countries get ahead and others do not.

Japan's Secret

IT IS NOT entirely clear why so little attention has been paid to national differences in intelligence. For something like 15 years now, there have been good data on the IQs of foreign populations, especially of the Japanese and other East Asians. The data plainly point to above-average IQs--i.e., above the American average in the Pacific Rim.

IQ comparisons between East Asians and other populations present one maddening difficulty: comparing verbal skills across cultures whose languages are so different. The researchers have handled this difficulty in a variety of ways. In some cases, their data are based on Chinese and Japanese versions of standard Western tests, especially the Wechsler Adult Intelligence Scale (WAIS) and the Wechsler Intelligence Scale for Children (WISC). Another solution has been to compare the two populations on tests like the Raven Progressive Matrices, a nonverbal IQ test that measures abstract reasoning ability.

In any case, the test results show not only a higher average in East Asia but a different structure of intelligence there: at any given IQ level, East Asians and Americans will display different patterns of strength and weakness. It appears that these differences cannot be explained entirely by social and cultural factors. It also appears that some of the particular strengths of the East Asians account for much of their pre-eminence in mathematics and science. So it seems reasonable to assume that the IQ data can help us answer one question that has long bedeviled American business and policymakers: How do the Japanese do it?

All five of the Pacific Rim countries--Japan, Korea, Taiwan, Hong Kong, and Singapore--have had outsize economic growth rates for three decades now. None of them has had a severe recession in that period, and all can point to staggering increases in living standards. Most American commentary on these countries has centered on their contributions to U.S. trade deficits. In this context, American executives increasingly bemoan the apparent superiority of the Japanese and other East Asian labor forces. Blaming the schools is very much the fashion in business these days. Why do Japanese children notoriously outperform their American counterparts by many different measures? The standard explanation is the one advanced in Winning the Brain Race, by David T. Kearns (chairman of Xerox) and Dennis P. Doyle (of the Hudson Institute), a book endorsed by then Secretary of Education Lauro F. Cavazos early in 1989. The explanation goes this way: "Japanese students do so well on international comparisons because they go to school for a longer day than Americans and a longer year ... A majority of American youngsters attend school 180 days a year, while the Japanese go for 240. The typical American student has an hour or less of homework; the Japanese student two and one-half hours or more."

There is little doubt that the Japanese educational system is extraordinarily effective, but it clearly does not explain most of the East-West performance gap. For one thing, other East Asian countries have school systems far less rigorous than the daunting Japanese model (and the Hong Kong schools are still Britishrun). Yet the record all across the Pacific Rim is one of superior academic achievement. In the recently completed international assessment of mathematics and science, 13-year-olds from nine different populations were studied by the U.S.-based Educational Testing Service. In math, the Koreans-the only East Asian group in the study-came in first (and the Americans last). In science, the Koreans came in second, a shade behind children from British Columbia. (The Americans were ninth.)

In any case, the superior talents of Japanese children cannot be entirely the product of Japanese schools, since some of the superiority is already observable in kindergarten. The February 1986 issue of Science carried a detailed report on an ambitious study led by Professor Harold W. Stevenson of the University of Michigan. A basic finding: American kindergardeners are about equal to the Taiwanese in mathematical understanding but both are far behind the Japanese. By the fifth grade, the Taiwanese have pretty much caught up with the Japanese and the Americans trail far behind.

East Is East and West Is West

THE FIRST reports of high East Asian IQs began appearing in the West in the late Seventies. Most of these reports were written by Richard Lynn, a British psychometrician based at the University of Ulster in Coleraine. Lynn, then 47, says he happened to read in 1976 that versions of the WAIS and WISC tests were becoming available in Japan. He wrote off for the Japanese test manuals, had them translated into English, and then began collecting and analyzing test results; before long he had a unique network of Far Eastern psychologists reporting regularly to Coleraine. More recently, he has begun acquiring test data from Mainland China.

These Chinese data are quite astonishing. Based on examinations using the Raven Progressive Matrices, the data show Mainland children with IQs averaging 101 on a scale where Americans are at 100. A one-point group differential would not ordinarily be worth mentioning. But IQs are powerfully correlated with several different indicators of material well-being. Scholars generally believe that something like 30 to 50 per cent of variability in IQ is attributable to environmental factors. And in this case, the one-point advantage is exhibited by a group living in abysmal poverty. The plain implication of those Raven scores is that Chinese IQs are potentially a fair amount higher than American IQs--also that a China liberated from Communism has a dazzling economic future.

One of Lynn's earliest reports dealt with a 1977 study in Singapore comparing the IQs of 13-year-olds in the Chinese and Malay populations. This study, also based on the Raven Progressive Matrices, showed the Chinese averaging 110, the Malays 96, on a scale where British children averaged 100. Lynn has also written in some detail about two 1982 studies of Chinese children in Hong Kong. One, based on Cattell's Culture Fair Test (it too emphasizes reasoning ability), showed a representative group of Chinese nine-year-olds averaging 106.7 in relation to a norm of 100 for American children. The other Hong Kong study looked at Chinese children of different ages and used the Raven plus a couple of other tests, including one vocabulary test. (A measure of verbal fluency and recall, it asked the children to write down the names of as many animals as they could think of in two minutes.) Lynn found that these data gave an average of 109 for the Chinese, against a mean of 100 for British children.

Lynn's most recent studies, which are heavily centered on Japanese test data, put a somewhat different spin on the basic message. In a 1987 paper, The Intelligence of the Mongoloids: A Psychometric, Evolutionary, and Neurological Theory, Lynn still says that these Pacific Rim populations have higher overall IQs than Westerners. But he now insists on a technical adjustment of the data that results in a gap of only three or four points.

Are We Getting Smarter?

THE ADJUSTMENT is worth a brief detour here, since it concerns a large issue to be addressed below-the issue of whether intelligence in developed countries has generally been rising over time. There is considerable division among scholars over this question, but Lynn is among those who believe that intelligence has been rising. Why would any such increase require an adjustment in the East-West IQ comparisons? Because the Asian tests were developed, standardized, and administered later than the tests on Americans and Europeans. If both groups were in fact getting steadily smarter, and if the East Asian test scores came later, then clearly some part of their reported advantage would be artificial.

Lynn's current thinking, in any case, puts far more weight on the differing structures of intelligence than on the IQ gap itself. In the 1987 paper, he writes: "The intelligence of the Mongoloids differs from that of Caucasoids more in the pattern or profile of their abilities than in the overall IQ." The model of intelligence used by Lynn in writing about structure is a familiar one to psychometricians. It features a three-layered hierarchy of abilities.

1. Sitting atop the hierarchy is "general intelligence," known as g. The basic thought behind g is that there is some general factor, affecting (though in different degrees) all kinds of mental abilities, from rote memory (weakly correlated with g) to abstract reasoning (practically synonymous with g). The alternative view is that intelligence is best conceived of as a cluster of discrete and essentially unrelated abilities.

The existence of g was once controversial, but today a substantial majority of scholars are committed to the idea of a general ability. Arthur Jensen of the University of California at Berkeley has argued - one might almost say crusaded - for g, and seems to have broadly prevailed. The basic reason for believing in some general factor in intelligence is that all those particular abilities correlate positively with one another, i.e., people who do above (or below) average in tests of any one ability are predicted also to do above (or below) average in any other. Using a mathematical technique called factor analysis, psychometricians can measure the extent to which various mental tasks are "g-loaded." Which is very close to saying the extent to which they rely on abstract reasoning.

2. The second layer of the hierarchy consists of two broad group factors, roughly characterized as verbal and visuospatial. A broad range of mathematical, scientific, mechanical, and engineering skills is powerfully influenced by the visuospatial abilities (e.g., the ability to imagine what an object would look like after being rotated in space).

3. In the third layer is a long list - it gets longer every year-of primary abilities. Some of these are thought of as particular forms of the broad verbal and visuospatial talents, but some others, like a memory for musical rhythms, are viewed as separate.

Back to the East-West differences, beginning at the top of the hierarchy. Lynn's first finding is that Japanese children are distinctly superior in g. At least, they are superior beginning around age six. At earlier ages, they fall short in g, apparently because they mature more slowly than Caucasian children. But after catching up, they pull steadily ahead; and on a scale where average American white children score 100, the Japanese score around 104 in the years between ten and sixteen.

In the second layer of the hierarchy, the picture is mixed. In the verbal factor, Japanese children are no standouts. They again start out behind their American counterparts in the very early years, but this time they do not catch up until around age nine or ten; furthermore, they do not then proceed to pull ahead of American children, but instead continue to score around 100 on average. Lynn believes that they are actually less talented verbally than American children. Many of the verbal skills can be raised by intensive study, so the apparent parity after age nine probably reflects the superiority of the Japanese educational system.

In the visuospatial factor Japanese superiority is discernible as early as age four and a half. With white Americans again normed at 100, the Japanese score around 105 in the years between ten and sixteen.

In the third layer, the specific skills break down about as you might expect. The Japanese are relatively weak on most verbal tests, but they are generally outstanding on tests measuring drawing ability, speed of perception, and thinking about objects being moved around in space. (In case you are wondering, they do about as well as Americans on musical rhythm.) Lynn finds that "These characteristics are also present among Mongoloids in the United States and, so far as the evidence goes, in the three Far East nations of Hong Kong, Singapore, and Taiwan."

The general rule, then, is that the East Asians excel at visuospatial tasks but are much weaker in the verbal sphere. Footnote to the rule: Lynn's Mongoloid-Caucasoid differences are somewhat analogous to (although greater than) male-female differences in Western societies- where men on average do far better than women in math and science, while women generally do better in verbal skills. Further footnote: American men seem about as good at the visuospatial tasks as East Asian women.

At Their Mother's Knee?

WHERE DID these East-West differences come from? Since so many of the ability differences are discernible in very young children, it is just about impossible to serve up a purely environmental answer to that question. Lynn himself believes in a mostly genetic answer, and has proposed an imaginative evolutionary explanation.

Severely compressed, his explanation goes about like this: Some sixty thousand years ago, when the lee Age descended on the Northern Hemisphere, the Mongoloid populations faced uniquely hostile "selection pressure" for greater intelligence. Northeast Asia during the Ice Age was the coldest part of the world inhabited by man. Survival required major advances in hunting skills. Lynn's 1987 paper refers to "the ability to isolate slight variations in visual stimulation from a relatively featureless landscape, such as the movement of a white Arctic hare against a background of snow and ice; to recall visual landmarks on long hunting expeditions away from home and to develop a good spatial map of an extensive terrain." These, Lynn believes, were the pressures that ultimately produced the world's best visuospatial abilities.

What about the relative weakness of Mongoloid verbal skills? In approaching this question, Lynn notes the familiar male-female ability differences. Like most scholars, he believes that these arose from evolutionary pressures in which men needed visuospatial skills for hunting, and women needed verbal skills for childrearing. The Mongoloids, Lynn believes, took this process a step further. Verbal skills are, of course, centered in the brain's left hemisphere. Lynn hypothesizes that among Mongoloids, the cortex in the left hemisphere was invaded and made to take on more visuospatial processing. There is some evidence to support this view of a different neurological structure for East Asians, although the case is admittedly not closed.

The verbal-visuospatial difference is not the only one that distinguishes Caucasoid and Mongoloid IQ patterns. The two populations also differ in the variability of their scores. A representative sample of Americans or Europeans will show more variability than will an East Asian sample. In the familiar bell- shaped distribution curve, the bell is much narrower for the Japanese- -which is what you would expect from such a homogeneous population.

This difference is a major matter, and it is worth focusing hard on the data. Just about all Western populations report a standard deviation of 15 IQ points. (The SD, a basic measure of variability, quantifies the extent to which a series of figures deviates from its mean.) But the SD for the Japanese and other East Asian populations appears to be a shade under 13 IQ points. That difference does not sound like a big deal, and, in fact, it does not change things much in the center of the distribution. But the smaller standard deviation, combined with the higher average IQ, has one important implication for the quality of the Japanese labor force. It means that Japan has relatively few low-IQ workers. In the U.S., about 25 per cent of all workers have IQs below 90 u. only about 15 per cent in Japan). For IQs below 80, the comparable figures would be around 10 per cent for the U.S., around 3 per cent for Japan. There are good reasons to believe that these differences are related to Japan's advantages on the factory floor.

It is curious that economists have been so slow to pick up on the link between IQ and economic performance. Indeed, it is especially curious given economists' increased interest in "human capital." In measuring the value of such capital, the economics profession never seems to get beyond education and training. In a brilliant essay published in 1983 (in a volume called Intelligence and National Achievement), Barbara Lerner chided them for being blind to the much superior measures provided by IQ and other test scores. Dr. Lerner, a psychologist and lawyer, was herself one of the first to write about the connection between IQ and national economies.

Predicting Success

THIS CONNECTION can be approached from several directions, but let us begin with this hard fact: People with high IQs tend to do well in life. There is, or should be, no real controversy about these statements: On average, the rich are more intelligent than the middle class. And the middle class are more intelligent than the poor.

To be sure, you will always pick up a certain amount of static about what's causing what. Do people become well-to-do because they are smart? Or are they smart because they started off with the kinds of advantages that well-to-do people pass along to their children? The answer is that a) the causation flows in both directions but b) the former effect seems much more powerful. The main reason smart people are well off is that modern market economies tend to reward people roughly in proportion to their productive skills, and these skills are closely linked to IQ.

If you pick a boy at random out of the teenage population, your best predictor of his adult "success" is not his family background (as measured, say, by his father's occupational status) but his present level of intelligence (as measured by his IQ). Correlation between a boy's teenage IQ and his adult occupational status: a potent 0.65. Between his father's occupational status and his own at middle age: a modest 0.35 or 0.40. (Corresponding figures for girls are, alas, hard to come by.)

The dollar value of IQ differences has been pointed up in a sizable number of studies. A Census Bureau study of veterans tested in 1964, when the men were in their early thirties, showed that a 15-point IQ difference had translated into an 11 per cent earnings difference. One famous study examined the careers of brothers who grew up together in Kalamazoo, Michigan, and had similar educational levels. (Kalamazoo, which has kept records of public-school IQs since 1928, has been a treasure trove for generations of sociologists.) Brother studies are illuminating because the boys have common family backgrounds but dissimilar IQs; on average, brothers' IQs differ by 11 or 12 points. The Kalamazoo data show that a 15-point IQ difference in the sixth grade was associated with a 14 per cent difference in annual earnings during the period between ages 35 and 59.

In talking with acquaintances about IQs and careers, I note that many people instantly gravitate toward a distinction. They accept that many prestigious and high-paying occupations-judge, research scientist, senior executive-must obviously be filled by people of superior mental ability; but they do not accept at all that IQ has anything to do with what they see as the ordinary run of jobs. My friends are not alone. Many industrial psychologists, many sociologists, and most corporate personnel directors will tell you that intelligence is nice but not necessary.

The conventional wisdom about this matter says that in filling employment openings, what you mainly need to know is the applicant's knowledge of the job, which you can typically infer from his experience, training, and formal education. You might also be interested in certain psychological traits of the applicant, like honesty and motivation; however, you have no reason to care about his intelligence, since the worker will absorb everything he needs to know through experience. An extreme form of this view was a quite serious suggestion put forward by sociologist Randall Collins in The Credential Society. Arguing that anybody could become a doctor, Collins proposed that we create our doctors by letting young people start out as hospital orderlies and gradually work their way up the medical career ladder.

The durability of the intelligence-doesn't-matter perspective is somewhat amazing. There is massive evidence demonstrating that in fact most occupations are filled within identifiable IQ ranges, and indicating that in most occupations, the best predictor of performance is not job experience or formal education but intelligence, especially intelligence defined as g. For generating some of these studies and pointing me to others, I am indebted to Robert A. Gordon of Johns Hopkins University and Linda S. Gottfredson of the University of Delaware, who together run the Project for the Study of Intelligence and Society.

The studies relating particular occupations to specified IQ ranges were developed principally by Professor Gottfredson and lean hard on data generated by the U.S. Employment Service with its so-called General Aptitude Test Battery. Although the USES is not exactly eager to emphasize the point, the GATB is a kind of disguised intelligence test. Herewith some of the IQ ranges (equivalent to scores on the WAIS) that Professor Gottfredson came up with based on analysis of the GATB data: physician, engineer: 114 or more; highschool teacher, real-estate sales agent: 108 to 134; fire fighter, police officer, electrician: 91 to 117; truck driver, meat cutter: 86 to 112. You will note that the IQ ranges associated with each occupation are fairly wide. However, the great bulk of jobs are filled near the midpoint of the ranges, so it is reasonable to think of each occupation as associated with a particular layer in the IQ pyramid. Or, to state the same thought in a possibly more provocative way, the occupational hierarchy is in large measure an intelligence hierarchy.

Why Test?

THE CASE for g as a predictor of job performance has been buttressed by an avalanche of empirical studies, centered on three huge data bases. One was put together by the late E. E. Ghiselli, a psychologist who was based at the University of California at Berkeley and spent a quarter-century (1949-73) of his professional life collecting and analyzing job-aptitude test records. A second data base consists of 515 studies performed over the years by the U.S. Employment Service. The third set of studies has been generated by the U.S. armed forces in connection with their never-ending efforts to ascertain which kinds of warriors should be assigned to which kinds of tasks.

In studies of these data bases, intelligence-test scores did best at predicting performance in jobs that had a high measure of complexity. But the

predictions were still useful even for relatively unskilled jobs. Writing in the December 1986 Journal of Vocational Behavior, industrial psychologist Frank Hunter of Michigan State (a major participant in the studies) reported on the positive correlation between test scores and measures of job performance. For high- complexity jobs, the correlation was 0.58. (Positive correlations range from 0 to 1.) For the lowest-level tasks, the figure was 0.23, which is still "high enough to yield considerable utility," Hunter wrote. He added that there were no jobs at all for which tests of intelligence were not useful predictors. To be sure, there are some low-complexity jobs for which they are less useful than tests of psychomotor ability.

Next question: How much do these predictions matter? Alternative form of the question: Are potential workers in any given job different enough in ability to justify a lot of testing before time is spent training them? The evidence suggests that they are. Or so you would conclude from a series of studies performed mostly by Hunter and industrial psychologist Frank L. Schmidt of the University of Iowa. They found that the differences between workers (as captured in objective measures of performance) were greatest in the more complex jobs but were significant at every level.

In a cluster of jobs that they ranked as "low complexity" (unskilled and semiskilled blue-collar jobs), the top 1 per cent of workers were about 50 per cent more productive than the average worker and three times more productive than the bottom 1 per cent. In "medium complexity" occupations (technicians and supervisors, for example), the top 1 per cent were 85 per cent above the average and 12 times better than the bottom 1 per cent. In the "high complexity" area (managers, professionals, and some technical workers), the top 1 per cent were 127 per cent better than the average; statistical complications made it impossible to quantify results for the lowest 1 per cent.

These data have some large and obvious morals. One is that companies able to identify good workers in advance should be in a position to do a lot for their productivity. This is precisely the claim being made for g testing: that at every level of complexity, it outperforms both "credentialism" and tests designed to measure specific job knowledge.

Another moral of the data is that you get the biggest payoff of all when you make superior predictions about high-complexity jobs. Hunter and Schmidt estimate that the productivity gain associated with a shift to g testing would be about two and a half times greater in high-complexity than in low-complexity jobs.

How much might the U. S. economy as a whole benefit from a broad shift to testing?. A U.S. Employment Service team led by Hunter had a swing at that question several years ago and came up with an estimate of nationwide productivity gains worth around $80 billion (in 1980 dollars). This estimate was evaluated critically by a National Research Council panel. Its scholars agreed that many individual companies would benefit from testing; however, they dissented from the view that the country as a whole could get much mileage out of it. In thinking about the national economy, they observed, you run up against the awkward fact that it is ultimately expected to hire just about everybody. And if the bad workers will end up getting hired anyway, the gains from universal testing will perforce be limited. The ultimate problem, said the Council, is that "use of the ... GATB will not improve the quality of the labor force as a whole."

The Council did not add: "and bring it up to Japanese standards."

Falling Scores

IF NATIONAL intelligence levels are a critical determinant of national performance, it obviously matters whether those levels are rising or falling. As previously hinted, the heavyweights of the psychometric profession are very much disagreed as to the direction of intelligence trends.

Begin with the fact that, by certain obvious measures, intelligence in America would seem to be slipping. Despite a partial recovery in the last few years, average scores on the Scholastic Aptitude Test are still well below the average of the late Sixties. There is broad agreement that much of the test-score decline can be explained by the broadening of the testee base-but also agreement that some part of the decline reflects a genuine loss in academic ability. And it is not only the broad average that shows slippage. College Board data also show a pattern of persistent decline in the fraction of testees who score at the higher levels. In 1990, for example, only 7.4 per cent scored over 600 on the verbal test (down from 11.3 per cent in 1972). SAT scores correlate heavily (about 0.8) with IQ test scores, so these data seem to signal an IQ decline.

To be sure, the SATs and other college admission tests are given to an unrepresentative group of young Americans-the half or so of high- school students who are college-bound. But the data for lower-school children are also depressing. A persistent theme in recent reports from the National Assessment of Educational Progress is the decline in children's thinking ability. Summarizing NAEP data that spanned two decades, the Educational Testing Service wrote last fall: "The NAEP results indicate a remarkable consistency across subject areas- students are learning facts and skills, but few show the capacity for complex reasoning and problem solving."

There was no shortage of explanations for the decline. The boob tube was an obvious candidate. Deterioration in the schools was another. Also available were some solid-looking "dysgenic" reasons, i.e., explanations based on data showing above-average birth rates in the low-IQ population. Since it is now widely accepted that intelligence is highly heritable-with scholars typically estimating that 50 to 70 per cent of variability in IQ is attributable to genetic factors- dysgenic data would afford a convincing explanation of declines in thinking.

But this brings us to a rather awesome anomaly: IQ scores are in fact rising. The data show that in recent years American children have rather consistently scored higher on IQ tests from the Thirties and Forties than did the children who originally took them. The test results imply gains of perhaps 0.3 IQ points per year, which translates into some 15 points in fifty years. Which further implies that the difference between us and our grandparents is roughly equivalent to the current black-white difference in America.

Such IQ gains are by no means unique to America. They are being recorded in developed countries everywhere and turn up not only in broad-based IQ tests like the WISC but also in sharply focused tests of reasoning ability like the Raven Progressive Matrices. Furthermore, the gains have not been confined to young children. In the Netherlands, Belgium, and Norway, there were huge gains for military recruits (contrasted with recruits of a generation earlier), demonstrating that whatever is happening has a certain amount of staying power.

The data are puzzling, to put it mildly. If intelligence were in some broad secular rise of the kind showing up in IQ scores, we would expect to see a world far richer in geniuses, colleges bursting with young prodigies, lower schools with far higher proportions of gifted children. Instead we have explosive demands for remedial reading courses, dismal SAT scores, and those disappointing NAEP findings. How can it be?

Several solutions have been proposed. James R. Flynn, a New Zealand academic who did much to collect and publicize the data on secular gains, believes the gains mean mainly that IQ tests are not worth much as gauges of intelligence. But this "solution" only leads to other difficulties-a major one being that the tests continue to retain their predictive power. Which has to mean that they are measuring something, whether or not it is called intelligence," that incorporates mental abilities in demand in developed countries.

Another approach has been to argue that intergenerational differences are meaningless: that intelligence is necessarily culture-bound and we therefore prove nothing by administering IQ tests from the past to children growing up in the Nineties. The import of this position is that each generation has its own learning, reasoning, and problem- solving challenges, and that it would be absurd to view Thomas Jefferson as unintelligent simply because a magically resurrected Jefferson would have trouble with much of the Wechsler Adult Intelligence Scale. "IQ simply cannot be used to compare individuals who have grown up in different eras," says Charles Locurto, a psychologist based at Holy Cross.

Richard Lynn has weighed in with a different answer. As noted above, he believes that the IQ gains are real and that intelligence levels are genuinely rising. Lynn believes the rise has been masked by academic declines reflecting poorer teaching and, perhaps, students less motivated to learn. He also believes that the IQ rise is attributable mainly to improved nutrition.

Like many of the academic scholars, I found his evidence on nutrition at once surprising and compelling. In addition, his hypothesis about the secular IQ rise seems to account for today's non-existent multitudes of prodigies. Lynn believes that earlier generations included much higher proportions of uneducated, malnourished, and generally retarded individuals with IQs around 60 or 70, while the individuals who accomplished anything were a relatively small elite. Today, Lynn believes, the IQ distribution is much less variable, reflecting the fact that nutrition-based gains have disproportionately benefited the low-scoring population-which would explain why the rising average IQ has not translated into an abundance of genius.

Cont'd
http://www.encyclopedia.com/doc/1G1-10650145.html