free web hosting | free hosting | Business Hosting Services | Free Website Submission | shopping cart | php hosting
DANIEL GOLEMAN'S EMOTIONAL INTELLIGENCE: WHY IT CAN MATTER MORE THAN IQ (1995)
  • False facts are highly injurious to the progress of science, for they often long endure. But false views, if supported by some evidence, do little harm.
  • (Charles Darwin, The Descent of Man)
    No "serious" book published in the 1990s received higher praise from the press than Daniel Goleman's Emotional Intelligence, and since its publication, in 1995, its influence has constantly grown. One need read no further than its pages xi-xii, where Goleman announces his purpose and introduces his key term, to discover the reason for its popularity:
  • a challenge to those who subscribe to a narrow view of intelligence, arguing that IQ is a genetic given that cannot be changed by life experience, and that our destiny in life is fixed by these aptitudes. This argument ignores the more challenging question: What can we change that will help our children fare better in life? What factors are at play, for example, when people of high IQ flounder and those of modest IQ do surprisingly well? I would argue that the difference quite often lies in the abilities called here emotional intelligence. And these skills, as we shall see, can be taught to children. [Goleman's italics]
  • Moreover, Emotional Intelligence has become the flagship of an armada of books and audiocassettes that Goleman and members of his immediate family have authored or co-authored. On August 3, 1998, Amazon.com listed 29, the large majority published after Emotional Intelligence; and their list is far from complete. The day before I looked it up, I read in the London Sunday Times about a book by Goleman that Amazon.com does not have, Emotional Intelligence in the Workplace. (The Times treated it as an authoritative study and discussed the impact it is having in the business world.)
    On October 2, 1998 the Johannesburg Star devoted a full-page (11) article, "EQ's Competitive Edge", to a new book by Goleman, Working with Emotional Intelligence. It noted that since the publication of Emotional Intelligence, "bookshelves have been flooded after numerous management gurus and authors translated the concept into the language of business people. Towards the end of 1997, EQ became the latest fad in business literature." The article then cited "Stephanie Vermuelen, director of Applied EQ Corporation" that "present research shows emotional intelligence to be a far greater predictor of personal success than its poor cousin, IQ, or level of education. She believes that businesses are wrongly pouring their money into a bottomless pit of skills training." The article also quoted other EQ experts, "who have developed EQ courses, which are now popular in the business world", that IQ may be related to as little as 4% of real-life intelligence."

    In the United States, Emotional Intelligence and its offspring will also provide much needed ammunition to two groups of people who recently have been thrown onto the defensive by powerful attacks. One is supporters of affirmative action. They have always opposed using achievement tests, academic grades and especially intelligence tests as criteria for hiring, promoting and university admissions because (non-Jewish, non-Oriental) minorities do much worse on them than whites do. They are now reeling from the first open, concerted attack they have ever had to face. For them Emotional Intelligence is a godsend, with its mission of liberating us from "the tests that tyrannized us achievement tests SATs [which] are based on a limited notion of intelligence, one out of touch with the true range of skills and abilities that matter for life" (Goleman p.38) Now they can do away with these "barriers to minority advancement" while pretending to abandon preferential treatment.

    The other group is professors of education, whose existence depends on the requirement that prospective public school teachers must pass their mind-numbingly-idiotic courses. These requirements increased greatly between the early 1960s, when less than one-quarter of American public school teachers had postgraduate degrees and 15% did not have Bachelor's degrees, and 1981, when half of American public school teachers had Master's degrees and over 99% had Bachelor's degrees. Probably not coincidentally, the average Verbal SAT score of high school students choosing education as their intended college major fell from 418 in the academic year 1972-73 to 389 in 1979-80.

    The main effect of requiring education courses is to keep intelligent, academically oriented people out of teaching. Goleman and his supporters would claim that the decline in SAT scores of education majors is nothing to worry about. However, numerous, extensive studies have shown no positive correlation between student performance and educational expenditure, student motivation, post-high-school educational intention, self-esteem or any other factor except one. That is a robust correlation between student performance and their teachers' scores on qualifying examinations that measure ability to read, write, solve mathematical problems, identify main ideas and sequential steps, draw inferences, etc. In other words, examinations that test intelligence, and whose scores correlate with their SAT scores.

    Some education courses are diluted versions of academic courses and some are about the history, philosophy, etc. of education. However, there are only two possible justifications for requiring prospective teachers to take education courses instead of courses that increase their knowledge of the subjects they will teach. One is that these courses help them to teach better. That may be true for one or two education courses. But retaining the present number of required education courses and even increasing them can be justified only by the argument that education should be largely non-academic, teaching children sensitivity to their own and other people's feelings, self-esteem, self-awareness, self-confidence, appreciation of differences, etc. Despite the opposition of parents, who want their children to be taught history, mathematics, literature, science, foreign languages, etc., by 1994 only 41% of the average American school day was spent on academic subjects and the average American high school student spent much less than half the number of hours studying academic subjects than the average French or Japanese high school student and less than a third of the German average (Sykes 1995, p.16). (T. Sowell (1993, p.290) observed that the education of more than forty million American school children is being sacrificed to the careers of less than forty thousand professors of education.)

    The orientation that Goleman champions will put professors of education back on the offensive. He anticipates opposition from their two archenemies (p. 280):

  • These programs are a major change in any curriculum. It would be naive not to anticipate hurdles in getting such programs into the schools. Many parents feel that the topic itself is too personal. Teachers may be reluctant to yield yet another part of the school day to topics that seem so unrelated to the academic basics. All [teachers] will need special training.
  • Ominously, among the praises that are quoted on the back of the dust jacket of Emotional Intelligence is one by the Chancellor of the New York City Board of Education. (In 1981 the average SAT score of New York State high school seniors was seven points above the national average; by 1995 it was nineteen points below. In 1995 NY State spent 56% more than the national average on education per student. Between 1981 and 1995 its expenditure per student per year increased from less than $4000 to $9300. In 1994 less than one-third of the education budget of NY City was for classroom education. A great deal of the rest was for psychologists, guidance councillors, etc. many of whom are specialists in emotional education. In fact, in 1991, 47% of the staff of all American schools were not teachers. (Sykes 1995, p.228; Economist, June 22, 1996, p.58))
    Below are excerpts from some reviews of Emotional Intelligence. The first is from the New York Times, by far the most read American newspaper among the people who shape American society (members of Congress, corporate executives, heads of major labor unions, etc.) and even more importantly among the producers and writers of television news programs, which are the major source of news for most Americans (Lichter et al.1986, pp.11-12). It also exerts a tremendous influence on the major news organizations in deciding which issues merit attention (Lynch 1989, pp.96-7, 108n3). It told its readers in a review entitled "Research Affirms Power of Positive Thinking" (Book Review, September 17, 1995, p.23):
  • Daniel Goleman beginning with a masterly overview of recent research in psychology and neuroscience make[s] lively connections between the wealth of new understanding and the riches of older wisdom Mr. Goleman believes we can cultivate emotional intelligence and improve the general life performance of the many children who now suffer because of our society's unbalanced emphasis on the intellectual at the expense of the affective dimension of personality. In his final section, he offers a plan for schooling to restore our badly neglected "emotional literacy". Mr. Goleman integrates a vast amount of material in an original and persuasive way.
  • The (London) Times Educational Supplement, probably the most highly regarded and influential periodical on education in the world, said (February 9, 1996, p.12),
  • IQ reflect[s] one's suitability to cope with a degree course But personality, temperament, character, drive, call it what you will, have as much if not more impact on success in life. Daniel Goleman's populist, but sensibly documented, American bestseller [is] a text for our times. The contents of Emotional Intelligence are considerable and important. Goleman has achieved an admirable synthesis of neural circuitry, psychological theory and popular common sense Goleman creates a powerful case for re-structuring what goes on in the classroom.
  • Time is probably the most influential general-interest magazine in the English language, maybe in any language. In its October 16, 1995 issue it had a glowing six-page review (pp. 68-75).
  • Cognitive theory could simply not explain the questions we wonder about most: why some people just seem to have a gift for living well; why the smartest kid in the class will probably not end up the richest Daniel Goleman has brought together a decade's worth of behavioral research ... He sees practical applications everywhere for how companies should decide who to hire [to] how schools should teach From kindergartens to business schools to corporations across the U.S., people are taking seriously the idea that a little more time spent on the "touchy-feely" skills so often derided may in fact pay rich dividends. Nowhere is the discussion of emotional intelligence more pressing than in American schools Many school administrators are completely rethinking the weight they have given to traditional lessons and standardized tests. Peter Relic, president of the National Association of Independent Schools, would like to junk the SAT completely [since it causes] an immense loss of human potential because we've defined success too narrowly.
  • Another influential periodical is Library Journal because libraries rely on it to decide which books to buy and which to display prominently. It reviewed Emotional Intelligence on page 194 of its September 1, 1995 issue:
  • The book calls for universal adoption of educational curricula that teach youngsters how to regulate their emotional responses and to resolve conflicts peacefully. Goleman summarizes much of the best psychological work of the last few decades Based on good empirical data this fine example [sic] is recommended for academic and larger public libraries.
  • Moreover, adulatory reviews are not confined to newspapers and magazines with aspirations to be regarded as intellectual. My last excerpts are from the two South African women's magazines with the largest circulation. Fair Lady told its readers (March 6, 1996, pp. 107-9), "It is clear that IQ offers little to explain the different destinies of people Yet our schools and our culture fixate on academic ability." Cosmopolitan (March, 1995, pp.55-7) began its review: "You might be a brilliant student but this won't get you to the top. What you need is not a high IQ, but a high score on a different kind of scale: an emotional scale."
    These reviews, and others like them, are crucially important. Subsequent readers made the logical assumption that they were written by people who know something about this subject and examined Goleman's claims carefully. Consequently, they are predisposed to accept what Goleman says. So his influence spreads through a ripple effect, as his ideas are absorbed by more and more people who do not read his books but are informed, second and third hand, that its ideas have been proved and are what experts think.

    No one denies that emotional and social abilities are vital for social, emotional and occupational success. What Goleman does not tell his readers is that since the 1960s, and with great intensity since the middle 1980s, psychometricians have developed, tested, validated and refined tests that measure a wide range of social and emotional abilities (and virtues). Some of these abilities are as obviously important as leadership, social insight, sense of responsibility, recognizing one's own emotions and inferring other people's emotions, honesty and altruism. Some are as seemingly trivial as appropriate dinner behavior. Scores on all these tests correlate very closely with scores on IQ tests and SATs, as do teachers' assessments of emotional and social strengths. (Brand 1996, p. 318 provides an extensive bibliography of studies going back to 1962; see also Legree 1995, Mayer and Geher 1996, Mussen et al. 1970, p. 174 and Jensen 1980, p.476.) So, "on 11 of 12 measures of social and emotional adjustment, gifted [as determined by high IQ] children in Grade 3 were found to be more advanced than average children in Grade 6" (Brand 1996, p.318). Another example is that when the WAIS IQ test was given to 185 architects, mathematicians, scientists and engineers who were rated by their colleagues as being particularly creative, their average IQ was in the highest two percent of the general population (Jensen 1980, pp. 355-6).

    The findings outlined above are well known and nearly universally accepted. Goleman shows that he is aware of them on page 44:

  • People with a high IQ but low emotional intelligence (or low IQ and high emotional intelligence) are, despite the stereotypes, relatively rare. Indeed, there is a slight correlation between IQ and some aspects of emotional intelligence. There is ample research on [testing] each of its [emotional intelligence's] components.
  • It is typical of Goleman, as I will show, not to cite any source for the "slight correlation' between IQ and emotional intelligence and to cite only an unpublished manuscript for the differences between high IQ and emotionally intelligent people that he outlines in the rest of page 44 and the top of page 45. However, his assertion about a slight correlation must be based on an acceptance of the validity of the many tests that have been devised to measure aspects of emotional intelligence. The only conclusion that can be drawn from the fact that Goleman did not try to refute the massive, uniform evidence for a very high correlation between scores on these tests and IQ is that he could not. Instead, he relied on the lie that the correlation is slight and assumed that the reviewers of his book would be extremely partial towards his views, totally ignorant about its subject and too lazy to find out even the most basic facts about it.
    However, Goleman was careless enough to put great emphasis on a study that should have shown even the most ignorant reviewer that his denigration of IQ tests is untenable. He states several times that self-control is the most important aspect of emotional intelligence (e.g. pp. 27, 56, 285). The title of the chapter he devotes to it is "The Master Aptitude" (pp. 78-95). There he says (p.81), "There is perhaps no psychological skill more fundamental than resisting impulse. It is the root of all emotional self-control". (Despite the "perhaps", Goleman never says anything like this about any other ability.) Goleman makes this assertion while describing what he clearly regards and expects his readers to regard as an extremely important experiment, which he calls "The Marshmallow Test" (pp.80-83). In the early 1960s an experimenter put one marshmallow in front of each of a group of four-year old children. He told them that he was going out on an errand and that on his return he would give another marshmallow to those who had not eaten the marshmallow in front of them. Goleman says,
  • The diagnostic power of how this moment of impulse was handled became clear some twelve to fourteen years later The emotional and social difference between the grab-a-marshmallow preschoolers and their gratification-delaying peers was dramatic. Those who had resisted temptation at four were now, as adolescents, more socially competent: personally effective, self-assertive, and better able to cope with the frustrations of life. They were less likely to go to pieces, freeze, or regress under stress, or become rattled or disorganized when pressured; they embraced challenges and pursued them instead of giving up even in the face of difficulties; they were self-reliant and confident, trustworthy and dependable; and they took initiative and plunged into projects.
  • The third or so who grabbed for the marshmallows, however, in adolescence were more likely to be seen as shying away from social contacts; to be stubborn and indecisive; to be easily upset by frustrations; to think of themselves as "bad" or unworthy; to regress or become immobilized by stress; to be mistrustful and resentful prone to jealousy and envy; to overreact to irritation with a sharp temper
    When the tested children were evaluated again as they were finishing high school, those who had waited patiently at four were far superior as students [Goleman's italics] to those who acted on whim. They were better able to put their ideas into words, to use and respond to reason, to concentrate, to make plans and follow through on them, and more eager to learn.
    Goleman ends the subchapter on this experiment by stating that it "underscores the role of emotional intelligence as a meta-ability, determining how well or how poorly people use their other mental capacities."
    There is one more difference between those who waited for two marshmallows and those who did not, a difference to which Goleman assigns great importance:
  • Most astonishingly, they [the marshmallow-resisters] had dramatically higher scores on their SAT tests. The third of children who at four grabbed for the marshmallow most eagerly had an average verbal score of 524 and quantitative (or "math") score of 528; the third who waited longest had average scores of 610 and 652, respectively - a 210 point difference in total score.
  • Anyone familiar with SAT scores knows that these are extremely large differences and that even the lower average score is much higher than it would be in a random sample.Before 1995, when SAT scores were renormed, if all eighteen-year olds took them, the average verbal score would be 325 and the upper one percent would begin at 569 (Herrnstein and Murray 1994, p.767n.3). (These are not the distributions for those who took the SAT but for all 18-year olds.) The reason that the average scores in the marshmallow test were so high is that these children were an extremely select group: mostly children of faculty members and graduate students at Stanford University. Clearly if the marshmallow test were tried on a random sample of the population the range of scores, and consequently the difference between the marshmallow grabbers and resisters, would be much greater. (Goleman says they were children of faculty, graduate students "and other employees" at Stanford, but his source (Shoda et al., 1990, p.980) says only faculty and students.)
    Despite denials from the Educational Testing Service, which administers SATs, anyone who knows anything about them knows that they are intelligence tests and scores on them correlate extremely closely with IQ (Jensen 1985, p.203; Snyderman and Rothman 1988, pp.151-2; Fallows 1982, p.42; Herrnstein and Murray 1994, p. 38). Goleman himself says, "the SAT [is] highly correlated with IQ" (p.86) and "SAT scores are" "a surrogate" "for IQ" (p.315, n.15). So how does Goleman use these SAT scores to attack IQ?
  • At age four, how children do on this test of delay of gratification is twice as powerful a predictor of what their SAT scores will be as is IQ at age four; IQ becomes a stronger predictor of SAT only after children learn to read. This suggests that the ability to delay gratification contributes powerfully to intellectual potential quite apart from IQ itself.
  • The source Goleman cites for the correlation between IQ at age four and SAT scores is "a personal communication from Phil Peake". I have read many scholarly books and articles and have never seen an unpublished (and therefore uncheckable) communication used even as corroborating evidence, let alone the only cited evidence for a crucial argument; but, as the reader will see, Goleman uses this type of source often. In fact, the degree of correlation between IQ before the age of six and at eighteen (when IQ3DSAT) is debated (e.g. Carroll 1997, p.45; Brody 1992, p. 233; Levin, p. 103).
    However, it would be irrelevant if IQ at the age of four has no correlation with later IQ. Goleman grants that there is a significant correlation between IQ at four and eighteen (half the correlation of the Marshmallow Test). But Herrnstein and Murray (1994, pp.53, 130), who argue (and prove) that IQ from the age of 7 or 8 is an amazingly accurate predictor of occupational, social and emotional success, state that there is none: "Up to about 4 or 5 years of age, measures of IQ are not much use in predicting later IQ" (but page 243). This question is not important because the uses of IQ tests that Goleman attacks begin after the first grade; and it is certain that a person's IQ at the age of 10 correlates very closely with his SAT score and that the odds are 2 to 1 that an adult's IQ will be within three points of his IQ at eight (Fallows 1982, p.42: Herrnstein and Murray 1994, pp. 396, 590-91; Levin 1997, p.62; Neisser et al. 1996, p. 81).

    Goleman's point is that "ability to delay gratification contributes powerfully to intellectual potential quite apart from IQ itself". This assertion is clearly nonsensical. IQ measures certain abilities. It does not contribute to those abilities any more than a thermometer contributes to heat. (In parenthesis Goleman cites a study that shows that poor impulse control predicts delinquency better than IQ. He discusses this study on pages 236-7. I will analyze it later.)

    As I said, Goleman attaches great importance to the marshmallow test, and with good reason. It is by far the clearest, most cogent proof of his thesis that he provides. On his showing, what he calls "the master aptitude", self-control, and every other social and emotional characteristic of any value are measured with great accuracy by SATs, which he knows are a form of IQ test. These emotional and social qualities are also predicted by the marshmallow test, but that can be used only at one age. Moreover, it can differentiate only between those who take a marshmallow and those who do not and it does not give any indication of different types of abilities (e.g. verbal and quantitative). SATs and IQ tests do what the marshmallow test does and everything it does not. They indicate many levels of many types of abilities. Also, SATs take a short time and are extremely easy to administer, requiring only one person to watch a room of forty or fifty students. SATs and IQ tests are, in fact, among the most brilliant and useful inventions of all time.

    The reviewers of Emotional Intelligence saw how important the marshmallow test was to its thesis. For example, the review in the Times Educational Supplement is entitled "Cleverness Is Two Marshmallows". The beginning of the review in Time is, "It turns out that a scientist can see the future by watching four-year-olds interact with a marshmallow." Time then summarizes the experiment and the differences in adolescence between those who took the marshmallow and those who did not, including, "when students in the two groups took the SAT the kids who had held out longer scored an average of 210 points higher". The reviewer concludes her description of this clearly crucial experiment with what she thinks is its greatest significance: "And it [i.e. all these vital characteristics that this experiment predicts] doesn't show up on an IQ test." (In fact, Goleman wrote that at the age of four IQ has half the predictive power of the Marshmallow Test and its predictive power increases greatly after six. Time changed that, thus making the experiment fit Goleman's thesis perfectly.) However, at the end of the Time review, the reviewer quotes with obvious approval the recommendation of P. Relic, president of the National Association of Independent Schools, "to junk the SAT completely. 'Yes, it may cost a heck of a lot more money to assess someone's EQ [i.e. emotional quotient] rather than using a machine-scored test to measure IQ,' he says." Here, correctly, the SAT is called an IQ test and the ease with which it is marked is noted. If the reviewer had read what she wrote at the beginning of her review, she would know that SATs are an excellent measure of EQ.

    (The president of the National Association of Independent Schools is hardly the impartial expert that the Time review assumes. "One of the major factors in breaking the near-monopoly of private preparatory schools in supplying students to the elite colleges was the development of a nationwide, standardized, college entrance test [i.e., the SAT]" (Sowell 1993, p.127). Moreover, before 1919 no American university ever considered using any non-academic criteria for admission. "Character", geographic distribution, being the son of an alumnus, etc. were introduced to save elite universities from being inundated by poor, immigrant Jews and keep their student-bodies predominantly upper-class, Anglo-Saxon and private-school educated. (Steinberg 1974, pp. 5, 9, 19-31; 1981, 237, 248; Oren 1985, 42-3, 46-58; Synnott 1979, pp. 76-7, 92, 106-8, 112, 155))

    In containing a blatant self-contradiction, the Time review imitated one of Emotional Intelligence's most pervasive and obvious characteristics. On page 38, in a chapter entitled "When Smart is Dumb", Goleman attacks the tyranny of

  • the achievement tests [and] SATs that determined what, if any, college we would be allowed to attend - [and] are based on a limited notion of intelligence, one out of touch with the true range of skills and abilities that matter over and above IQ.
  • He expresses similar views frequently:
  • High school valedictorians and salutatorians get excellent grades in college but are not particularly successful professionally later. So "achievement as measured by grades tells nothing about how they react to life". (pp. 35-6)
  • "Social intelligence is both distinct from academic abilities and a key part of what makes people do well in life". (p.42)
    "The independence of emotional from academic intelligence [there] is little or no relationship between grades or IQ and people's emotional well-being". (p.57)

    "In keeping with findings about other elements of emotional intelligence, there was only an incidental relationship between scores on this measure of empathetic acuity and SAT or IQ scores or school achievement tests." (p.97)

    However, Goleman more frequently asserts that emotional health is crucial for attaining high marks in school and college and on achievement and intelligence tests. That means that marks in school and college and on achievement and intelligence tests should be accurate indicators of a person's emotional health, and for that reason, if no other, be excellent predictors of personal and occupational success and extremely reliable criteria for hiring and promoting employees and admitting applicants to universities. For example, on page 193 Goleman adduces "a growing body of evidence showing that success in school depends to a surprising extent on emotional characteristics". He mentions only one piece of evidence, the all-important marshmallow test. "As we saw in Chapter 6, for example, the ability of four-year-olds to control the impulse to grab for a marshmallow predicted a 210-point advantage in their SAT scores fourteen years later." When Goleman outlined that experiment (p.82) he italicized that the marshmallow-resisters were "far superior as students" on a wide range of academic abilities. Whenever Goleman extols a specific emotion he asserts that it is crucial for academic success:
  • "Anxiety also sabotages academic performance of all kinds. The more prone to worries a person is, the poorer their academic performance, no matter how measured - grades on tests, grade point average, or achievement tests". (pp. 83-4)
  • "Good moods mak[e] it easier to find solutions for problems, whether intellectual or emotional". (p.85)
    "As with hope, optimism predicts academic success" (pp. 86-8). (This is part of a demonstration of the vital role optimism and hope play in occupational success. On page 88, as opposed to the statements in this paragraph and pp. 35-6 (quoted above about high school valedictorians), Goleman says that high school grades are poor predictors of college grades.)

    "Students who get into flow [a crucial virtue in Emotional Intelligence] do better as measured by achievement tests". (p. 93).

    "Another cost to these children is doing poorly in school; depression interferes with memory and concentration". (p. 243)

    "Dropping out of school is a particular risk for children who are social rejects. [Their] dropout rate is between two and eight times greater than for children who have friends. Two kinds of emotional proclivities lead children to end up as social outcastes" (p. 250).

    Moreover, during his discussion of the vital role optimism plays in occupational success (p.89, n.23), Goleman cites an article he wrote in the New York Times (February 3, 1987, p. C3), in which he quotes Dr. Martin Seligman, one of the heroes of Emotional Intelligence, that "pessimistic children also do less well on [academic] achievement tests".
    I did not check Goleman's references for these assertions. If they are based on reliable data, they help to explain the massive evidence, some of which is presented in The Bell Curve, that there is an extremely close correlation between success in school, university, work, marriage, parenthood and good citizenship and that IQ tests, which were designed to measure academic ability, are also excellent indicators and predictors of social abilities and emotional health. Unfortunately, Goleman's sources are probably no more reliable for his assertions that academic success requires emotional health than are the sources he cites for his assertions that the opposite is true, which I did check and outline below.

    Incredibly, the reviewers completely missed Goleman's constant assertions that emotional health determines performance in school and on academic tests, and therefore school performance and academic tests must be extremely reliable measures of emotional health.

    I now will analyze Goleman's attack on intelligence tests and academic performance as predictors of success in non-academic activities. At stake is not only the use of these measures as criteria for hiring and university admissions but also the nature of education. Goleman never denies that IQ/SAT scores are genetically determined and he sometimes explicitly assumes that they are. For instance, when he introduces his thesis, in pages xi to xii, he states that the existence of emotional skills is the reason why

  • people of high IQ flounder and those of modest IQ do surprisingly well ... And these skills can be taught to children, giving them a better chance to use whatever intellectual potential the genetic lottery may have given them.
  • So if IQ tests measures emotional health, then Goleman's testimony contributes to the already irrefutable evidence that emotional health is as hereditary as intelligence. That means that no matter how important emotional abilities are, they cannot be taught. Consequently, education should concentrate on what can be taught and what no one knows innately: science, literature etc. (The ability to learn these subjects is innate, but not the subjects themselves.)
    In fact, Goleman has known for a long time that personality traits are mostly genetically determined. An article he wrote on the front page of the Science Section (Section C) of the New York Times of December 2, 1986 was entitled "Major Personality Study Finds That Traits are Mostly Inherited". Its first sentence is, "The genetic makeup of a child is a stronger influence on personality than child rearing". Goleman was reporting on the first publications by Thomas Bouchard Jr. and his team at the University of Minnesota of their study of twins, especially monozygotic twins raised apart. All their subsequent findings strengthened those in their initial publications. Goleman kept from his readers the fact, which he has long known, that monozygotic twins who are raised from infancy by different families in different social environments and who attend different types of schools develop the same personalities and emotional strengths and weaknesses, a fact that is crucially important in assessing his arguments and recommendations.

    Goleman's attack on the significance of IQ is also based largely on omission of crucial evidence with which he was familiar. A year before the publication of Emotional Intelligence, The Bell Curve created a tremendous uproar. Anyone who reads any magazines or newspapers or watches any television had to be aware of its existence. Its main thesis is that IQ tests predict not only occupational success with remarkable accuracy but also whether a person will be a criminal, get married, divorced, have an illegitimate baby, be a good mother, etc.; in other words, emotional intelligence. The Bell Curve bases this thesis on meta-analyses of a total of over a thousand studies of civilian employment, databases that were carefully compiled by the US Defence Department of the military careers of cumulatively nearly a million people, and an extremely intensive ongoing study conducted on over 12,000 people for twelve years.

    It must be emphasized that The Bell Curve's main purpose is to present this massive evidence for the predictive accuracy of IQ tests. The issue of hereditability is secondary; and the few serious reviewers who questioned Herrnstein and Murray's arguments for the hereditability of intelligence still acknowledged the manifest incontrovertibility of their demonstration of IQs predictive accuracy. One example is C. Finn (Commentary, January 1995, pp. 76-80): "there can be little or no doubt about their findings on the predictive power of IQ in relation to success and failure in contemporary U.S. society". Another example is the black scholar Thomas Sowell in the American Spectator (February 2, 1995, pp.32-3). (Because Sowell questioned the hereditability of intelligence, this review was reprinted in an anthology of attacks on The Bell Curve, entitled The Bell Curve Wars, edited by S. Fraser. The passage below is on page 71.)

  • Herrnstein and Murray establish their basic case that intelligence test scores are highly correlated with important social phenomena from academic success to infant mortality, which is far higher among babies whose mothers are in the bottom quarter of the IQ distribution. Empirical data from a wide variety of sources establish that even differing educational backgrounds or socioeconomic levels of families in which individuals were raised are not as good predictors of future income, academic success, job performance ratings, or even divorce rates, as IQ scores are. Even in non-intellectual occupations, pen-and-pencil tests of general mental ability produce higher correlations with future job performance than do "practical" tests of the particular skills involved in those jobs. In terms of logic and evidence, the predictive validity of mental tests is the issue least open to debate. On this question, Murray and Herrnstein are most clearly and completely correct.
  • Obviously if Goleman wanted his thesis to be taken seriously, he had to prove that there are major defects in Herrnstein and Murray's evidence and/or that his own evidence is more cogent. However, Goleman mentioned The Bell Curve only twice. On page 34 he quoted a statement from it that the relationship between IQ and making a million dollars or becoming a senator is weak. No one has ever thought that being successful in politics requires intelligence or emotional health; but Herrnstein and Murray marshal massive, unequivocal evidence that the abilities measured by IQ tests are necessary to become rich, although some people with high IQs prefer to use those abilities to become scientists, professors, etc. (See also Murray 1997.) This passage is an example of the tendency of proponents of the hereditarian position to make unsubstantiated statements that contradict their own evidence. I provide an example later.
    Goleman cited The Bell Curve again on page 80 for what he obviously regarded as crucial evidence in support of his thesis.
  • The added payoff for life success from motivation, apart from other innate abilities [italics added], can be seen in the remarkable performance of Asian students in American schools and professions. One thorough review of the evidence [he cites The Bell Curve] suggests that Asian-American children may have an average IQ advantage over whites of just two or three points. Yet on the basis of the professions, such as law [italics added] and medicine, that many Asians end up in, as a group they behave as though their IQ were much higher The reason seems to be that from the earliest years of school, Asian children work harder than whites. Sanford Dorenbusch [wrote], "While most American parents are willing to accept a child's weak areas and emphasize the strengths, for Asians, the attitude is that if you're not doing well, the answer is to study later at night, and if you still don't do well, to get up and study earlier in the morning. They believe that anyone can do well in school with the right effort." In short, a strong cultural [italics added] work ethic translates into higher motivation, zeal, and persistence - an emotional edge.
  • This passage is typical of Emotional Intelligence in that Goleman provides no reference for Dorenbusch's assertions, his argument is clearly wrong and his ignorance is mind-boggling. His ignorance is manifested especially by his use of law as an example of superior Oriental achievement. Anyone who knows anything about intelligence tests knows that the extent of the Oriental-Caucasian difference in intelligence is debated; but in every study that has ever been done, Orientals and Caucasians manifest different strengths. Orientals do much better than Caucasians on the subtests of IQ tests that measure nonverbal (especially spatial) intelligence and on intelligence tests, like Raven's Progressive Matrices, that depend partially on spatial intelligence. (Contrary to what some psychometricians think, RPMs are loaded on the spatial factor: Carroll 1997, p.35.) But Orientals do worse than Caucasians on subtests of IQ tests that measure verbal intelligence. For example, Asians score about a third of a standard deviation higher than whites on the Math section of the SAT and a third of a standard deviation lower on the Verbal. (Lynn and Song 1994; Vernon 1982, pp. 75,123-4, 273; Herrnstein and Murray 1994, pp.273-4; Jensen 1982, p.134; Zajonc and Mullallay 1997, p. 696)
    As a result, in the USA architecture is the profession in which Chinese and Japanese are most overrepresented in relation to their numbers, whereas there are less than a quarter as many Chinese-American lawyers as their proportion in the American population despite the fact that Orientals receive preferential treatment in admission to law schools. (In 1990-91, 1.55 more Asian-Americans were admitted to American law schools than would have been admitted if they were whites with the same qualifications: Thernstrom 1998A, p. 17) (Exact statistics are not available for other American Orientals, but all are definitely underrepresented among lawyers.) Also congruent with their performance on IQ tests is that the number of American Orientals who write non-technical books - fiction, non-fiction, poetry, drama, etc. - is less than half of their proportion in the American population and in 1996 only seven Orientals received PhDs in History in the United States. (Vernon 1982, p.179; Weyl 1989, pp. 21-2,58-60,71-2; Herrnstein and Murray 1994, p. 455-6; Henderson et al. 1998, p. 41)

    It is worth noting that Jewish mental strengths are the opposite of Orientals'. Wherever Jews live, they attain extra-ordinarily high scores on subtests of verbal intelligence and low scores on subtests of spatial intelligence (Storfer 1990, pp. 314-23). So architecture is the only profession in which Jews are underrepresented. Conversely, there are seven times more Jewish lawyers in the United States than Jews in the general population and fifteen times more Jews on the faculty of elite law schools than their proportion of the American population. Similarly, over five times as many Jews write books that are reviewed frequently than their proportion of the American population. (The proportion of both Jews and Orientals in American Men and Women of Science, Frontier Science and Technology and Who's Who in Engineering is many times higher than their proportion of the American population.) (The average Jewish IQ is higher than the average Oriental IQ.) (Weyl, 1989, pp.21-2, 49-52, 58-60, 71-2, 77; Mayer 1967, pp.96-7; Farber and Sherry 1997, p. 58; MacDonald 1994, pp. 128-9)

    The main source Goleman cites for the difference between Oriental and Caucasian intelligence, The Bell Curve, like every other book or article that discusses this subject, emphasizes that the difference between Oriental verbal and nonverbal intelligence is much greater than the difference between Orientals and Caucasians in overall intelligence. An example is on page 300: "This finding [invariable Oriental superiority in nonverbal over verbal IQ] has an echo in the United States, where Asian-American students abound in engineering, in medical schools, and in the sciences, but are scarce in law schools and the humanities and social sciences."

    The other source Goleman cites is J. Flynn's Asian-American Achievement beyond IQ. Flynn is one of the very few scholars who argues that Orientals do not have higher average IQs than Caucasians. He has a political purpose, to show that the reason American blacks are less successful than whites is not genetic, and his arguments are easily refuted (Lynn 1993, 1996). However, Flynn, like everyone else who writes about this subject, points out that Orientals have much higher nonverbal than verbal intelligence and he frequently cites N. Weyl's The Geography of American Achievement, which provides detailed statistics on Oriental over-representation in non-verbal professions and under-representation in law and other verbally oriented professions.

    So if Goleman knew anything about intelligence tests, he would know that the explanation he quotes for superior Oriental achievement (Oriental parents, as opposed to white parents, do not accept their children's weaknesses and emphasize their strengths) is not only impossible but risible. In fact, the profile of Oriental achievement is a clear and irrefutable illustration that IQ tests and SATs are remarkably accurate predictors of career success and that cultural and emotional factors, like "a strong cultural work ethic" and "motivation, zeal, and persistence" are totally irrelevant. If these or other cultural and emotional factors were important, there would be as many, or at least nearly as many, Oriental lawyers, historians and novelists as there are Oriental engineers, architects and physicists.

    It is true that American Orientals are, on average, significantly richer, more successful occupationally, better educated and less likely to get divorced or commit crimes than white Americans (Eastland 1966, p.171; Wilson and Herrnstein 1985, pp. 471-2). However, anyone who read The Bell Curve would know that these differences are to be expected if IQ predicts these factors and if Orientals average three IQ points higher than whites, because on pages 364 to 368 Herrnstein and Murray provide a detailed description of the immense differences between two groups of people whose average IQs are three points different from each other. They also point out there that in a bell curve distribution, like that of IQ scores, small differences in averages produce great differences in the curve's tails. If population A has an average IQ of 100 and population B of 97, then 31% more of the former have IQs over 120 than the latter and 42% more over 135. (Herrnstein and Murray 1994, p. 276 estimate that the difference between whites and Orientals is three IQ points, not "two or three points", which is what Goleman says they estimate.)

    I said above that Goleman's argument from Oriental achievement is typical of Emotional Intelligence in three ways: it contains an unsubstantiated assertion, it is manifestly wrong and it displays incredible ignorance. It is also typical in a fourth way. It contains an obvious self-contradiction (motivation is innate and cultural) and it blatantly contradicts fundamentally important views that Goleman champions elsewhere. On pages 44 to 45 Goleman says that high-IQ-low-emotional-intelligence people are "ambitious and productive", which is clearly meant to be regarded as insufficient. On pages 35 to 36 he outlines a study which showed that high school valedictorians and salutatorians got excellent grades in college but were not particularly successful professionally ten years later. The reason is that their academic achievements indicated that they are merely "the 'dutiful' - the people who know how to achieve in the system". Yet, according to Goleman, "the remarkable performance of Asian[s] in [the] professions" is caused by their extraordinary dutifulness. (Typically, Goleman cites only one source for the characteristics of high-IQ people and one for the study of valedictorians, and these are an unpublished manuscript and a newspaper article. With considerable difficulty I managed to find the latter. It is clearly unreliable, as I will show.)

    On pages 94-5, in the same chapter in which Goleman provides the above-quoted explanation of Oriental professional success, he quotes with approval Howard Gardner, who is the hero of Emotional Intelligence:

  • Howard Gardner sees flow, and the positive states that typify it, as part of the healthiest way to teach children, motivating them from inside "We should use kid's positive states to draw them into learning in the domains where they can develop competencies," Gardner proposed to me. The strategy revolves around identifying a child's profile of natural competencies and playing to the strengths A child who is naturally talented in music for example, will enter flow more easily in that domain Knowing a child's profile can help a teacher fine-tune the way a topic is presented and offer lessons at the level - from remedial and advanced - that is most likely to provide an optimal challenge.
  • This is the opposite of the way Goleman says Orientals obtain their "remarkable" success. However, if Goleman knew anything about this subject, he could have offered Orientals as an excellent illustration of the ideal of guiding education by the principle of flow. This principle is clearly based on the premise that each person's types and levels of abilities are genetically determined and unalterable ("natural") and that educators must recognize this as the factor that controls what and how much each student should be taught. Wherever in the world Orientals live, no matter what the socioeconomic conditions in which they were raised or the nature of the school system they attended or whether they had been adopted in infancy by white parents, they invariably score much higher on nonverbal intelligence tests and nonverbal subtests of IQ tests, especially those that measure spatial intelligence, than on those that measure verbal intelligence. These scores on IQ tests are paralleled by the professions they invariably enter and in which they excel. (Lynn and Song 1994; Frydman and Lynn 1989; Vernon 1982, pp. 75,123-4, 273; Herrnstein and Murray 1994, pp.273-4; Jensen 1981, p.134)
    As Goleman observes and is obvious, this "strategy [of teaching by flow] revolves around identifying a child's profile of natural competencies" and "knowing the level - from remedial to advanced -" at which to teach each student. The example of Orientals, which he offers, is an excellent illustration that IQ tests are ideal instruments for identifying which subjects a student should study, to what level (elementary school, high school, college, graduate or professional school) he should pursue them, what type of college or graduate school he should attend and which career he should enter.

    Clearly not one reviewer of Emotional Intelligence read it with even slight care, or knew anything at all about this subject or was willing to spend a day glancing through The Bell Curve. The reviewer for Time was worse than ignorant and lazy. She informed its readers that "among the ingredients for success, researchers now generally agree that IQ accounts for about 20%". No other statement in the Time review could have more influenced the 98% of its readers who have no knowledge of this subject besides what the media tell them. Time's review is by someone named Nancy Gibbs, and at its end four people are listed as having reported it. Its readers undoubtedly assumed that when its editors chose five people to report a subject to which they devote a large part of an issue and whose great importance they emphasize, they would have taken care that at least one of them had a slight knowledge of the subject or would take a day or two to acquire it. Consequently, they must have assumed that when they are told that this is what researchers agree, that must be true: Goleman's views are mainstream, the only questions are ones of degree and pro-IQ advocates are a small group of eccentrics.

    When I read the Time review I wondered what its source could be for IQ accounting for only 20% of success. Its context clearly refers to occupational success. Herrnstein and Murray (1994, p.74) point out that the lowest estimate ever arrived at by a meta-analysis for the accuracy of IQ and IQ-like tests predicting work productivity is 35-40%, and the explicit purpose of the panel that made this estimate was to warn against using intelligence tests to exclude blacks from employment. James Heckman (1995, pp.1095, 1107) gives a somewhat higher estimate than that in a scurrilous attack on The Bell Curve, which is marred by misrepresentations (Murray 1995, pp.20-21). In fact, Heckman concedes (pp. 1098-9) that Herrnstein and Murray "cite numerous scholarly studies that refute critics of IQ and aptitude tests. They demonstrate convincingly that psychometric tests predict productivity, even if not perfectly." (The last clause shows how desperate Heckman was to find a criticism. No one maintains that IQ tests predict anything perfectly.)

    However, Time was talking not about success at performing a job, but success as measured by the status of one's job. (Time mentions social class and luck as being as important as IQ in determining success.) Accurate as g is in predicting performance in a job, it is much more accurate in predicting status of occupation; that is, there is an extremely close correlation between a child's score on g-loaded tests and his occupational status as an adult. Status of occupations has been determined by many surveys from 1920 until the present in the USA, Britain, the Netherlands and the Soviet Union. In some surveys people were asked to rate occupations on the basis of how much intelligence is needed to perform them; in other surveys on the basis of how much prestige they have; in other surveys on the basis of which are the most desirable. The results of these surveys "are amazingly consistent with one another and are highly stable throughout the industrialized world and from one decade to another". Not one survey has yielded results that are not highly congruent with the others. The overall rank-order correlation between the studies ranges from .95 to .98. The correlation between an occupation's rank and the rank order of average IQ of its members is .90-.95. The average IQ of the members of high-status professions has also remained remarkably constant over decades. For example, the average IQ of doctors has remained at about 125 for four decades. (Jensen 1980, pp.339-41; 1993, pp156-7; 1998, p. 292-3, Sloshberg and NesSmith 1983, p.160; Gordon 1997, p.204)

    So I was mystified where Time got a 20% correlation between IQ and occupational success, since 20% is only slightly more than half of the lowest correlation that has been estimated between IQ and job performance, and correlation between IQ and occupational status is higher than between IQ and job performance. But I read the Time review before I read Emotional Intelligence. When I read Emotional Intelligence I found Time's source. On page 34, during a discussion of occupational success, Goleman states, "at best, IQ contributes about 20% to the factors that determine life success". The Time reviewer took that figure and added to it "researchers now generally agree".

    The source Goleman cites is an article by Howard Gardner called "Cracking open the IQ Box", which is reprinted in an anthology of articles attacking The Bell Curve, entitled The Bell Curve Wars, edited by S. Fraser (1995, pp. 26-7). Gardner is a professor at Harvard's Graduate School of Education, and as I said, the hero of Emotional Intelligence. For example, on page 37 Goleman states, "If anyone sees the limits of the old ways of thinking about intelligence, it is Gardner." This homage is easy to understand. Gardner is the most influential academic in the world whose ideas parallel Goleman's. Goleman's admiration for Gardner is reciprocated. The dust jacket of Emotional Intelligence quotes the following praise by Gardner: "At last, a psychology that gives equal time to the intelligence of emotions. Never before have Dan Goleman's highly acclaimed gifts of writing been so effectively employed. A good and important book."

    In the article that Goleman cites, Gardner says (pp.26-7), "Nearly all the reported correlations between measured intelligence and societal outcomes explain at most 20 percent of the factors contributing to socioeconomic status [SES]". Gardner does not mention a source for a single one of these "reported correlations", even though he is attacking The Bell Curve, which never makes a statement on this subject, or any other, without discussing the extremely extensive and intensive studies on which it is based, every one of which yields much higher correlations. In fact, in the thirteen pages of Gardner's article, he cites not a single article and only two books, neither of which pertains to correlation between IQ and SES. Nor does he try to refute a single statement that Herrnstein and Murray make. Typically of critics of The Bell Curve, the only parts he mentions are those in which Herrnstein and Murray bend over backwards to contradict their own evidence and parts that he misrepresents. So he observes (p. 27) that Herrnstein and Murray

  • note that IQ has gone up consistently around the world in this century that when blacks move from rural southern to urban northern areas, their intelligence scores also rise; that black youngsters adopted in households of higher socioeconomic status demonstrate improved performance on aptitude and achievement tests; and that differences between the performances of black and white students have declined on tests ranging from the Scholastic Aptitude Test to the National Assessment of Educational Practice [sic]. Herrnstein and Murray say that the kind of direct verbal interaction between white middle-class parents and their preschool children "amounts to excellent training for intelligence tests".
  • I discuss the constant rise in IQ scores in "The Flynn Effect" Excursus and "The American Psychological Association" section of XXXXXX. The third and fourth statements so blatantly and totally distort what The Bell Curve says (pp. 309-10; 294-5; 722-3,notes 65-7) that they must be conscious lies. I could not locate the last statement. (Gardner supplies no page numbers from The Bell Curve.)
    The second statement is an example of the powerful tendency of hereditarians to undermine their own arguments and evidence. On page 303 Herrnstein and Murray outline a study reported by Jensen in 1977 (Developmental Psychology 13, pp.184-91) that in black families in rural Georgia older siblings have lower IQs than younger siblings, but there is no comparable difference between white siblings in this sample or between black siblings in Berkeley, California. They say this is proof "that environment can depress cognitive development". This study is emphasized by nearly every hereditarian as counter-evidence to their position. For example, Seligman (1992, pp.158-59) outlines it and draws from it the conclusion, "So environmental effects are real." Hereditarians emphasize it even in short articles. For instance, on page 33 of the New Republic, October 31, 1994 Murray and Herrnstein say, "There are, of course, many arguments against such a genetic explanation. Many studies have shown that the disadvantaged environment of some blacks has depressed test scores." Then they mention this study as if it were one of many that have shown this. In fact, it is the only study that has ever indicated environmental impact on blacks' IQ scores. They themselves say accurately in The Bell Curve (p. 303) that this is "the clearest evidence that the disadvantaged environment of some blacks has depressed their test scores". They also cite elsewhere in The Bell Curve Gordon's article, in which he points out (1984, pp. 366-71) that no other study ever had similar results, including other studies of black and white children in Georgia and in rural North Carolina, that the intelligence test used in this study, the CTMM, has been found to be unreliable on several occasions and consequently is not generally used by psychometricians and it clearly was totally unreliable in this study because the average IQ of the black children in it was 71 at the age of 12, declining to 65 at the age of 16. These are uniquely, in fact, bizarrely, low scores, even for blacks in the rural South.

    The other means Gardner uses to attack The Bell Curve are also typical of its critics: unsubstantiated assertions and references to unspecified studies whose results are described so vaguely and briefly that the reader has no way to verify them. However, there is one exception, one study that Gardner describes in enough detail for it to be evaluated (p.31-2):

  • To understand the effects of culture [rather than genetics], no study is more seminal than Harold Stevenson and James Stigler's book The Learning Gap: Why Our Schools Are Failing and What We Can Learn from Japanese and Chinese Education (1992). In an analysis that runs completely counter to The Bell Curve Stevenson and Stigler show why Chinese and Japanese students achieve so much more in school than do Americans. Genetics, heredity, and measured intelligence play no role here. East Asian students learn more and score better on just about every kind of measure because they attend school for more days, work harder in school and at home As a Japanese aphorism has it, "Fail with five hours of sleep, pass with four". Americans score near to last on almost all measures save one: When you ask Americans how they think they are doing, they profess more satisfaction than any other group. (Gardner's italics)
  • I pointed out that this explanation of Oriental success means that they manifest an exaggerated caricature of the "dutifulness" that Goleman claims accounts for high school valedictorians and salutatorians not being particularly successful professionally, and it is antithetical to the ideal of guiding education by "flow", for whose merits Goleman quotes Gardner. In fact, the Oriental model of effective education, which Goleman and Gardner both hold up as an ideal, is the exact opposite of the emotional and social education that they champion. According to supporters of traditional, academic education, like Thomas Sowell (1993, p.3), "nothing so captures what is wrong with American schools as" the studies that show that American students know much less than Eastern Asian students but think they know more. This demonstrates that progressive education, among whose highest goals is instilling in students a feeling of self-esteem, is "a success in its own terms - though not in any other terms". Moreover, as will be illustrated below, despite receiving an education that is diametrically opposite to the type that Goleman and Gardner champion, Japanese and Chinese have much lower rates of crime, divorce and other manifestations of emotional and social malfunction than Americans do. This is an excellent illustration of the irrelevance of emotionally oriented education for emotional health.
    As was stated, in his thirteen-page attack on The Bell Curve Gardner cites no articles and only two books. He cites one of these books, Lisbeth Schorr's Within our Reach, because he claims that it shows that there are social programs that have improved education, health care, family planning, etc. But these issues are only of the most marginal relevance to The Bell Curve. I could find in this book no mention of the validity of IQ tests or the hereditability of intelligence and only two references to the content of education, both of which are antithetical to what Goleman and Gardner champion. On pages 226-7 Schorr says that there is "wide agreement on the attributes that various researchers found crucial to making schools effective". The first is "an emphasis on academics". The others include "the importance of a coherent curriculum" and "protecting school time for learning". On page 243 she praises a program for putting "heavy emphasis on reading".

    So Gardner relies completely on Stevenson and Stigler's "seminal" study to refute The Bell Curve. This is bizarre. Herrnstein and Murray know that if two people have the same IQ and one works harder in school, is taught better or merely stays in school longer, he will learn more, up to an upper level determined by his innate intelligence. In fact, they say that (pp.419-34) and devote Chapter 18 of The Bell Curve to suggesting how to improve American education. Most of the criticisms they make there of American education and changes that they suggest are the same as those made by Stevenson and Stigler, whose book they praise and quote (p.437). Moreover, Stevenson and Stigler are concerned only with elementary education. Herrnstein and Murray know that nearly all Orientals and Caucasians and most Blacks have the ability to learn more at that level than they now do; but as the educational level rises, fewer and fewer people have the innate ability to profit from it, no matter how hard they try or how well they are taught. As Murray (1995, page 25) wrote,

  • Do we know how to take a set of youngsters with a given IQ and reliably improve their educational achievement? Yes. Do we know how to take a set of youngsters with a given tested IQ that would not allow them to become engineers, and reliably raise their cognitive functioning so that they can become engineers? No.
  • The relative intelligence of Orientals and Caucasians is irrelevant to Stevenson and Stigler's book and they mention it only in passing. It was in an article written in 1985 that they argued that Orientals do not have higher average IQs than whites. This was based on the only study ever done that obtained this result. Herrnstein and Murray (1994, pp.274-6, 287, 717n15) and Lynn (1997; 1993, p.238) have pointed out obvious defects in it. The most serious is that its American data is from the city of Minneapolis, whose white population has an average IQ several points higher than the average white American. Stevenson and Stigler have never answered these objections, nor does Gardner, in an article attacking The Bell Curve. (If someone used IQ tests done at Jewish schools as typical of Caucasians, he could prove that Caucasians have higher average IQs than Orientals.)
    I will briefly outline Stevenson and Stigler's book. The total lack of any arguments or evidence that support the anti-IQ, pro-emotional-intelligence case is well illustrated by the fact that its most distinguished academic supporter must rely on and praise fulsomely a book that attacks everything he (and Goleman) champions. Stevenson and Stigler point out that Chinese and Japanese elementary school children are superior to American children at every academic subject and they enjoy school much more, although American students and their parents rate their academic ability and achievement much higher than do Oriental students and their parents (pp.28,30-31,48,57,66-7,70,111,117-18,166). They use these facts to attack the emphasis in American education on building self-esteem and self-confidence, an emphasis that most proponents of progressive education, including Gardner in this article (pp.30-31) and Goleman (pp. 192-4, 86, 243-4), want increased, since they argue that high self-esteem is essential for effective learning.

    Another superiority of Oriental over American education that Stevenson and Stigler praise concerns the preparation of teachers.

  • The number of years spent in formal education [is] more than eighteen for the Americans we interviewed, compared to about fifteen for teachers in Sendai and Taipei. Some American teachers had master's degrees; none of the Asian teachers had received more than a bachelor's degree. In fact, some of the teachers in [Communist] China had no more than a high school education, and many of the teachers in Taiwan had only five years of schooling after grade nine. Asian teachers-to-be are more likely than Americans to major in liberal arts and to take courses in substantive disciplines - for instance, mathematics or literature - rather than in methods for teaching these subjects. American teachers-in-training generally major in education, and take many courses in teaching methods. (pp.158-9) (Remember that Gardner is a professor at Harvard's Graduate School of Education.)
  • Stevenson and Stigler praise Asian education not only for not wasting time and money on idiotic and useless education courses, but also for spending much less money on schools in general; for example, "China devotes 3.7% of its gross national product to education - a modest amount compared to the 6.8% of the much larger GNP of the United States." (p.133). Some of this saving is effected by school buildings in China and Japan not having central heating, libraries, gymnasiums, computer rooms or other facilities that Americans regard as necessities (pp.131-2). Another money-saving practice of Oriental education that Stevenson and Stigler praise is that many more students are in each class, from "thirty-eight to fifty children" (p.62).
    Stevenson and Stigler mention another wasteful expenditure of American education that Asian schools avoid (p.133). They have "no assistant teachers, school psychologists, counsellors, or social workers Is not the family, the Asian parent asks, responsible for handling children's emotional problems?" But Goleman's main thesis is that American schools do not devote enough time and attention to students' emotional education. However, Stevenson and Stigler report that children in Asian schools have fewer emotional disorders than American children and enjoy school more (pp.57,66-7,70). Moreover, the crime rate in Japan is infinitesimal compared with the United States, where much formal education is already concerned with emotional and social development. For example, the robbery rate is 120 times higher in the United States than in Japan (Wilson and Herrnstein 1985, p.453).

    Stevenson and Stigler attribute all the superiorities of Chinese and Japanese over American education largely to the adherence of each

  • with one of two basic positions. The first, "Intellectualism", holds that the goal of education is the mastery of core academic subjects by everyone. The opposing position regards the intellectualist agenda as old fashioned [It does] not include a primary emphasis on academic learning. The anti-intellectualist position has gradually dominated. Between 1910 and 1950 the proportion of academic subjects in American high school curricula fell by almost 60 percent. The old academic curriculum was virtually replaced by the so-called life-adjustment curriculum. Japanese and other Asian educators proceeded to develop educational systems in which students had to adapt to the unwavering standards of demanding academic curricula The goal of elementary education is unambiguous: to teach children academic skills and knowledge. Many Americans place a higher priority on life adjustment and the enhancement of self-esteem than on academic learning. They assume that positive self-esteem is a necessary precursor of competence. (pp.107-11) American teachers and the American public hold a notion of the ideal teacher that is very different from that held in Asia. (p.166; cf. 54-7))
  • Gardner is right that Stevenson and Stigler provide extremely powerful, indeed seemingly irrefutable evidence for how education should be conducted. It should concentrate completely on academic competence and knowledge and totally ignore emotional and social development. Consequently, supporters of that position cite Stevenson and Stigler extensively; for example, C. Sykes in Dumbing Down Our Kids: Why Our Children Feel Good About Themselves but Can't Read, Write or Add, pages 16-20,28-30,49,103,296-7. In fact, Gardner is the one American educational theorist whom Stevenson and Stigler name as epitomizing what they argue is wrong with American education (p.134).
    It is worth making explicit now something that was implicit in my introduction of Gardner. If there were any legitimate evidence or arguments to support the anti-IQ or pro-emotional-education case, Gardner would know them. In fact, if anyone would be able to fabricate plausible lies to support either position, it would be Gardner.

    (I mentioned that this article by Gardner is reprinted in an anthology of anti-Bell Curve articles, entitled The Bell Curve Wars. The article following Gardner's is by Richard Nisbett. He accuses Herrnstein and Murray of being "strangely selective" in their reports on the effects of childhood intervention in raising IQ. He wonders if they are "unaware of the very large literature that exists on the topic of early intervention". But from this "very large literature", Nisbett mentions only one study, reported in Pediatrics 1992, about a program that raised the IQs of low-birth-weight babies nine points at the age of three. Nisbett does not mention that interventions usually produce dramatic gains at first, which then disappear. Even more dishonest is that Nisbett does not mention that that happened in this study. He does not tell the reader that in the follow-up of this study, published two years later, the children involved had an advantage of only 2.5 points on one IQ measure and one-fifth of one point on another over the control group, who had no intervention. (Murray 1995A, p. 29; 1995, pp. 18, 25 (an exchange of letters between Nisbett and Murray); 1996, pp. 572-3) The article before Gardner's in The Bell Curve Wars is by Stephen Gould. I deal with that in an accompanying document.)

    I will now outline, as they occur, every book and article that Goleman cites as anti-IQ evidence and that I could locate. The Inter-Library Loan Division of my university could not find some, even though the academic and public libraries of southern Africa are well supplied with books and scholarly journals on psychology and related subjects. I looked for the articles and books that are not in southern Africa in the renowned New York Public Library on Fifth Avenue in Manhattan. Any book or journal I did not locate must be extremely marginal.

    When I read Emotional Intelligence my first reaction was amazement how few studies Goleman cites to substantiate his assertions and what small samples they involve. Anyone who is familiar with this subject at all would immediately realize that even if every one of those studies were conducted with impeccable rigor and Goleman reported their results accurately, the fact that he could find so few anti-IQ studies and that they all involved minute sample sizes would itself be nearly irrefutable proof that the pro-IQ position is unassailable. Clearly Goleman counted on the reviewers being totally ignorant of this topic and too lazy and/or prejudiced to spend even a day glancing through The Bell Curve.

    As I mentioned, if they had looked through The Bell Curve, they would have seen that the predictive power of IQ for personal and occupational success has been demonstrated by meta-analyses of a total of over a thousand studies of civilian employment, databases that were carefully compiled by the US Defence Department of the military careers of cumulatively nearly a million people, and an extremely intensive ongoing study conducted on over 12,000 people for twelve years. They would also have seen in The Bell Curve an emphatic warning that individual studies of small samples are unreliable and an explanation of why that is true (pp. 69-71). It is a basic statistical principle and obvious to common sense that the size of a database is an extremely important component of a study's reliability and that meta-analyses of many studies are much more reliable than individual studies. (Rushton 1995, pp.17-24 provides several excellent illustrations of this principle.) The most obvious reason is that distortion of sampling error is inversely related to sample size, but there are other, more complicated reasons also.

    On page 27 Goleman writes, "These deficits are not always tapped by IQ testing In one study, for example, primary school boys who had above average IQ scores, but nevertheless were doing poorly in school were found to have impaired frontal cortex functioning." Goleman refers to an article by P. Harden and R. Phil in Journal of Abnormal Psychology 104, 1995, pp. 94-103, entitled "Cognitive Function, Cardiovascular Reactivity, and Behavior in Boys at High Risk for Alcoholism". This article reported a study that added to the already massive evidence that alcoholism is hereditary. The authors sum it up on page 100, "The results suggest a relationship between SOMMAs [sons of male multigenerational alcoholics] and their performance on frontal lobe tests." They add, "However, the generalizability of this study is restricted by the small sample size". (The study involved only fourteen SOMMAs and fourteen normal boys as a control group.)

    On the same page and the next Goleman writes,

  • In a work of far-reaching implications for understanding mental life, Dr. Antonio Damasio has made careful studies of [people whose] decision-making is terribly flawed - and yet they show no deterioration at all in IQ or any cognitive ability.
  • After this passage Goleman continues to cite Damasio's book as an extremely authoritative source for his basic thesis (pp. 28,52-3). This book, Descartes' Error, is three hundred pages, excluding its index. The index has no entryfor "IQ", "intelligence" or any related term. On page xii Damasio says he has seen many cases of intelligent people who make disastrous mistakes, but he gives no references. I read the book twice and found only three mentions of intelligent people making calamitous decisions, one of whom made his disastrous mistake in 1848 (p.3 ff.). For only one of them does Damasio say he had a high IQ (p.40). There are no discussion of any evidence or reference to any studies of any of these cases, so the reader has no way of reading any more about them than what Damasio tells him. It is easy to see why this book is close to Goleman's heart.
    On page 35 Goleman states,
  • IQ offers little to explain the different destinies of people with roughly equal promises, schooling, and opportunity. When ninety-five Harvard students from the classes of the 1940s were followed into middle age, the men with the highest test scores in college were not particularly successful compared to their lower-scoring peers in terms of salary, productivity, or status in their field. Nor did they have the greatest life satisfaction, nor the most happiness with friendship, family and romantic relationships.
  • Ninety-five people is a tiny sample to oppose to the samples used in The Bell Curve, which involve tens and hundreds of thousands of people. Nevertheless, several reviewers of Emotional Intelligence mentioned this study as one of Goleman's more cogent proofs. However, the 396 pages of the book which Goleman cites for this study, Adaptation to Life by G. Vaillant, do not contain a single statement that resembles this. I read it through twice looking for something that even remotely approximates Goleman's claim. (It has no index.) Then I reread his reference to this book (p.314, n.4). I discovered that after citing it, Goleman says, "The average SAT score of the Harvard group was 584 Dr. Vaillant told me about the relatively poor predictive value of test scores for life success in this group of advantaged men." The reason I missed that statement when I first read this reference is that I have read many books and articles on this and other subjects and I have never seen a personal communication or other unpublished source cited as the sole evidence for an assertion; in fact, I have never seen such a source cited even as corroborating evidence. I became accustomed to them as I read more of Emotional Intelligence. This statement, told to Goleman nearly two decades after the book that recorded the experiment was published (1977), is what he opposes to the massive evidence, the many carefully controlled studies of huge numbers of people, that indicate that IQ is an extremely accurate predictor of success. Goleman does not even say that he saw a study indicating this, and the reader is supposed to take this seriously. Moreover, before SATs were renormed in 1995, if every eighteen-year-old took them, 584 would be in the upper one- percent of scores on the Verbal part (Herrnstein and Murray 1994, p.767, n.3). It is extremely improbable that ninety-five people in that rarefied range would contain subgroups large enough for their differences to be statistically significant. However, all ninety-five together might form a statistically significant sample of people, all of whom are characterized by extremely high SAT scores. So I will outline what this book, which Goleman recommends, says about a group of people whose average SAT score was in the upper one percent of the American population.
    The study involved is called the Grant Study, after the philanthropist who financed it. The subjects were all men who graduated from Harvard College between 1942 and 1944. They were chosen on the basis of academic achievement, being untroubled by physical or psychological disturbances and being independent (pp.30-32). Their socioeconomic background was not particularly privileged. Half of their parents were not college graduates and half paid a significant proportion of their educational expenses by working. "At the end of thirty years, the socioeconomic differences among the subjects upon graduation had no correlation with any of the outcome variables" (p.33, Vaillant's italics). Their most distinguishing characteristic was superior academic performance. Sixty-one percent were graduated cum laude or higher, as opposed to twenty-six percent of their classmates (p.33). (Remember that this is at Harvard.) This is significant since Goleman often contends that college grades are irrelevant to future success. However, the Grant men, who were characterized by extremely high SAT (3D IQ) scores and university grades, were much more successful than the average American by every measure of professional and personal success that Vaillant mentions. For example,
  • Most rose to the rank of officer and made distinguished records for themselves in the less academic atmosphere of World War II. There they were judged for skills other than intellectual achievement. Over ninety percent have founded stable families. Virtually all have achieved occupational distinction. (p. 4) The subjects have become bestselling novelists and cabinet members, scholars and captains of industry, physicians and teachers of the first rank, judges and newspaper editors (p.5). More often than not, the Grant men were the most occupationally successful of all their siblings (p.31). [In World War II] a third of the men were in sustained combat for ten days or more the men reported far fewer symptoms of nausea, incontinence, palpitations, tremor, and giddiness than other men under acute battle conditions. Only ten percent went into the army with commissions, but seventy-one percent were officers when discharged. (pp. 34-5) Many more [of the Grant men than their classmates] described their work as "extremely satisfying". At age forty-seven only eighteen percent were even twenty pounds over their optimal weight, and only thirteen percent averaged five days or more of sick leave a year. These figures are much lower than the general population. (pp.36-7) None was convicted for a crime. At an average age of forty-five, roughly eight percent of the Grant Study men were in Who's Who in America. By age fifty, twelve percent were in American Men of Science. (p.38)
  • Vaillant also compares the professional and personal success of the Grant men with the subjects of the Terman Study (pp.37-8). It would be strange if he did not. Few books on the subject of intelligence do not mention the Terman Study (e.g. The Bell Curve, p. 57 ("Terman's famous study")). Even Goleman must have known about it, if he read Vaillant's book, which he cites as an important source. That he makes no attempt to challenge it must indicate that he could not.
    The Terman Study "was one of the truly great social-science research projects of the twentieth century" (Seligman 1992, p. 44). Between 1921 and 1928, Lewis Terman of Stanford University selected 857 boys and 671 girls in California public elementary and high schools solely on the basis of their IQs. All had IQs of 135 or above, over 95% were above 140; their average IQ was 151 and their median IQ was 147. (An IQ of 135 is in the upper 1% of the general population, 140 in the upper 0.4%, 150 in upper 0.1%.) Terman then kept systematic records of their lives. The most complete report is Terman and Melita Oden's The Gifted Group at Midlife, published by Stanford University in 1959. The most recent report is Carole Holahan and Robert Sears' The Gifted Group in Later Maturity, published by Stanford University in 1995. By all occupational criteria, they were much more successful than most Americans. Among the men, 88% were in the highest status occupations: professional, proprietor, manager and executive, compared with 20% of men in the general American population. Not only was their average occupational level remarkably high, but they were much more successful than other men at the same occupational and educational level. Their average income was 61% higher than the average income of American professionals and managers, and the average income of those who had not gone beyond high school was comparable with those who graduated from college. Of the six with the highest incomes, only one was a college graduate. Incredibly, at an average age of forty-five, 12BD% were in Who's Who in America; and by the age of fifty, 21% were in American Men of Science. Among the women, about half were housewives, which was normal for the time; but one-ninth were high-level professionals, 8% were business executives and seven were in American Men of Science (which was renamed American Men and Women of Science in 1971). They even had a lower mortality rate and many fewer accidents than most Americans of their age.

    Immediately following his citation of the Grant Study, Goleman says,

  • A similar follow-up in middle age was done on 450 boys who grew up in a "blighted slum" a few blocks from Harvard. A third had IQs below 90. But again IQ had little relationship to how well they had done at work or in the rest of their lives To be sure, there was a general link (as there always is) between IQ and socioeconomic level at age forty-seven. But childhood abilities such as being able to handle frustrations, control emotions, and get on with other people made the greater difference.
  • Again, anyone at all familiar with this subject would realize that 450 people is a trivially small sample to oppose to the huge databases that show the opposite. In fact, that an opponent of IQ has to resort to such obviously insubstantial evidence is strong proof of the unassailability of the pro-IQ position. But, also again, even this evidence does not exist. Goleman distorted the source to which he refers (J. Felsman and G. Vaillant, "Resilient Children as Adults: A 40-Year Study", in pages 289-314 of The Invulnerable Child, edited by E. James and B. Cohler). On page 297 is a table of the correlations between scores on six measures of the subjects' childhood strengths and weaknesses and their socioeconomic status (SES) at age 47. These measures were Boyhood Competence (how well the subjects coped as boys with part-time work, household chores, sports, etc.), Childhood Environmental Strengths (frequency of childhood problems with physical, social and mental health; parental relationships conducive to developing trust, autonomy and initiative), Childhood Emotional Problems Subscale (childhood emotional problems, how "good-natured" and sociable the subjects were as children), an IQ test given to the subjects when they were children, Childhood Environmental Weakness (lack of family cohesion, being raised apart from parents, lack of paternal affection and supervision) and parental SES. Of these childhood variables it was IQ that correlated by far most closely with SES at 47: .35. (A correlation of 1.00 means that two entities are identical.) But that figure understates what the correlation would have been in a random sample because the range of scores for these subjects was smaller than for the general population. The authors of the study point out, "sampling bias included the exclusion of the severely delinquent [and] the intellectually gifted, and blacks and women" (p. 290). (On the page on which Goleman mentions this study he shows he is aware that correlations are decreased by a sample that is restricted to an attenuated range when he says that the Grant Study took place when the IQ spread at Ivy League colleges was wider than it now is.)
    On pages 290 and 292 of the article to which Goleman refers its authors mention another distorting factor: 61% of the subjects' families were foreign born, which probably diminished the reliability of the verbal subtests of the IQ test they took. For that reason, the only other table of correlations in the article (p.301) provides scores on the block design subtest of the IQ test that these boys took, since it is nonverbal and highly g-loaded (explained below). That table compares the thirteen of the subjects from multiproblem families who turned out the best and the thirteenth who turned out the worst. As in all studies of this type, parental SES had no effect; but the average IQ of the worst outcome group was 88 and of the best outcome group 101. That would be a very large difference for any sample. It is especially large for this group, all of whom were white males from the same neighborhood and extremes were omitted even from that group. But the difference between the best and worst outcome groups was even greater on the block design subtest of the IQ test they took: 11.2 and 7.5. Since the block design subtest is highly g-loaded (i.e. it correlates closely with scores on other types of intelligence tests) we must assume that if most of this sample had been raised by English-speaking parents, the difference in IQ between its most and least successful members would have been much greater than thirteen IQ points.

    I will now remind the reader that on page 34 of Emotional Intelligence Goleman states, "at best, IQ contributes about 20 percent to the factors that determine life success" and Time added "researchers now generally agree" to that assertion. Of Goleman's anti-IQ sources that I could find, Felsman and Vaillant's study, which I have been outlining, is the only one that provides a correlation between childhood IQ and adult SES. Without the serious sampling distortions that Felsman and Vaillant point out, their correlation between childhood IQ and adult SES would certainly be well over twice .20; although their estimate is much less than those of incomparably larger and more careful studies.

    Immediately after Goleman's assertion of 20% influence of IQ on success, he quotes the statement, "The vast majority of one's ultimate niche in society is determined by non-IQ factors, ranging from social class to luck." Time included that statement as part of what "researchers now generally agree" on. Goleman's quotation is from the article by Gardner that is his source for the .2 correlation between IQ and SES. Gardner (1995, pp.26-7) (who wrote "initial" social class) again provides not a single reference to support this claim, even though he is attacking The Bell Curve, which presents massive evidence to the contrary. However, of the studies that Goleman cites to denigrate the importance of IQ, every one that involves the effect of social background on success reports that that effect is slight or non-existent.

    On the rest of page 35 (and 36), Goleman mentions a study of high school valedictorians and salutatorians. They got excellent grades in college. "But by their late twenties they had climbed to only average levels of success". Goleman provides one reference for this study: "Karen Arnold, who did the study was quoted in The Chicago Tribune (May 29, 1992)." This is one of many times that the sole reference Goleman provides for a study is a newspaper article. Newspapers and magazines are legitimate sources for simple facts, like how much money was spent on education by a school system in a given year or average results on a specific test, but not for outlines of a study involving the relationship between two or more factors. In those very rare instances when a scholar does cite such a source as his only reference for a such a study, the reaction of other scholars is immediate and unforgiving; for example,

  • An example of [a] use of sources unsuited to a scholarly publication is his citation of The U.S. News and World Report An article in such a magazine cannot possibly include the subanalyses and collateral data which determine the meaning to say nothing of the environmental and historical conditions which initially differentiated the populations. (M. Deutsch, Harvard Educational Review 39, 1969, p. 525)
  • However, at least The U.S. News and World Report is available in many libraries. That is not true of the Chicago Tribune. But I did manage to locate this article. It reports what "an ongoing [i.e. unpublished] study is finding". The article's only source is what Karen Arnold and Terry Denny, who conducted the study, both of whom are professors of education, said they had found. Moreover, in the three and a half years between this article and the publication of Emotional Intelligence either Arnold and Denny published nothing in a scholarly journal on this subject or Goleman chose not to inform his readers about such a publication. However, even the article's summary of what Arnold and Denny claimed to have found clearly contradicts the conclusion that high school grades are unimportant. They found that of eighty-one valedictorians in 1981, who were in their late twenties at the time of the article, "just one-quarter were at the highest level of young professionals in their fields". In fact, one quarter is an extremely high proportion of any group of people to be at the highest level of young professionals. (Goleman misrepresents this. He wrote, "only one in four were at the highest level of young people in their chosen profession". A profession can refer to many types of work, but a professional is someone with a post-baccalaureate degree.)
    The Chicago Tribune article then observes, "the 46 women were doing much less well, by career standards, than the men, primarily because they placed greater importance on the family". From this it seems probable that nearly half of the men, who composed only 43% of the valedictorians, were in the highest level of young professionals.

    The greater number of female valedictorians raises an important consideration. It is well-known that teachers give higher marks to girls than to boys in subjects in which boys attain the same or higher marks on standardized tests. Many explanations have been offered: girls are better behaved, and/or more docile, and/or more pleasant, and/or better at routine aspects of subjects; for example, in mathematics, girls excel in computation (addition, multiplication, etc.), boys on tasks that require mathematical reasoning. (Klitgaard 1985, p. 91; Benbow 1988, pp.170, 173-4, 190, 198)

    (Arithmetic computation has little correlation with mathematical ability or general intelligence. Albert Einstein was not only one of greatest mathematicians who ever lived but he was also a brilliant writer and expert cello player, but he had great difficulty balancing his check book. Conversely, Mrs. Shakuntala Devi, who can multiply two 13-digit numbers in 28 seconds and compute the eighth root of a 14-digit number in ten seconds, has a normal IQ. (Coren 1994, p. 80; Seligman 1992, pp. 4-5))

    Furthermore, girls tend to take easier courses. As the Chicago Tribune article observes, an A in home economics is not the same as an A in calculus. The Chicago Tribune article also suggests the obvious fact that there is a tremendous difference among schools in the quality of their students and therefore among valedictorians from different schools. These are among the reasons that universities and employers have found it necessary to use standardized tests as well as school grades in selecting among applicants.

    So even this biased article shows that high school grades predict professional success and also shows why they must supplemented by standardized achievement and aptitude tests.

    From the bottom of page 36 until page 39 Goleman launches his most vitriolic attack on IQ tests and SATs. He cites only one piece of evidence to substantiate this assault: the lack of correlation between scores on the Stanford-Binet IQ test and the Spectrum test devised by Goleman's hero Howard Gardner. The source he cites is chapter 6 of Multiple Intelligences (1993), which Gardner edited. This chapter was written by Gardner and Mara Krechevsky. The subjects involved were four-year-old children. Gardner and Krechevsky state several times, "Given the limited scope of our sample population [less than 40 children], we are not prepared to draw general conclusions about four-year old children" (let alone anyone else) (p.93); "Of course, without a much larger sample, no firm conclusions can be drawn" (p.102); "Because of the small sample the study should be regarded as generating hypotheses rather than as conclusive in any sense" (p.105). Moreover, Goleman exaggerates the lack of correlation between the children's IQ scores and their performance on the Spectrum test. He says (p.39), "there was no significant relationship between the children's scores on the two tests". But Gardner and Krechevsky say that the three children who did the worst on the Spectrum test were among the five lowest in IQ, and the child who had the lowest IQ also did worst on the Spectrum test. (p.102).

    However, suppose the sample was large enough to be reliable and there was no correlation between scores on both tests; what would that indicate? Innumerable carefully conducted studies involving huge numbers of people have shown, without exception, that IQ scores are remarkably accurate predictors of academic, occupational, social and emotional success decades after the tests were taken. Gardner and Krechevsky's proof that scores on the Spectrum test are significant is that one year later some (but not all) of nineteen of the children involved exhibited the same strengths as they showed on the test. Even this absurdly trivial claim of predictive power is contradicted by their own analysis. For twelve of the nineteen subjects the criterion was that their parents or teachers said that they had the same strengths as they showed on the Spectrum test the year before (p.103). But four pages earlier Gardner and Krechevsky said that one of the greatest advantages of the Spectrum test is that it "identified twelve strengths that had not been identified by either parent or teacher" (italics in the original). (Gardner and Krechevsky also say (p. 106), "Of course, the Stanford-Binet Intelligence Scale [IQ test] is a standardized measure, with excellent internal consistency and high reliability. The measure is easily and efficiently administered.")

    On pages 86-9 Goleman states that optimism and a high level of hope are better predictors of college freshmen's grades than are SAT scores or high school grades. This is a truly stunning claim. For generations hundreds of the world's most brilliant psychometricians have studied the predictive power of all sorts of factors for university performance. These studies have often used databases of hundreds of thousands of people. The only question that remains is the relative predictive power of SATs and high school grades for different types of students and different types of colleges. Other information that is commonly used (letters of recommendation, biodata, essays written by students, interviews, etc.) have no or negligible predictive use for college grades. (Klitgaard 1985, p.108 supplies references to several reviews of large numbers of studies, all of which, without exception, came to this conclusion. Klitgaard, whom I cite several times, was Dean of Admissions at Harvard.) As Manning and Jackson observed (1984, pp.196-7),

  • It is doubtful that any other kind of test or even any other body of test validation research approaches the number of studies in which college admissions test scores are related to future academic performance. The studies have been repeated thousands of times, and the results quite consistently support the conclusion that the higher the test scores the more successful, on average, the students are in college and graduate study.
  • In fact, the SAT scores of even twelve and thirteen-year-olds are highly predictive of university performance, with regard to grades, awards and articles published (Benbow 1992).
    Moreover, as was observed, SATs are IQ tests and Goleman knows that. So it is worth adding that the accuracy with which IQ scores predict academic performance has been studied and re-studied since the first usable intelligence test was devised in 1905:
  • The Psychological Abstracts contain some 11,000 citations of studies on the relation of educational achievement to "IQ". If there is any unquestioned fact in applied psychometrics, it is that IQ tests have a high degree of predictive validity for many educational criteria, such as scores on scholastic achievement tests, school and college grades, retention in grade, school dropout, number of years of schooling, entering college, probability of receiving a bachelor's degree. (Jensen 1998, p.277)
  • What does Goleman oppose to this massive evidence, which is accepted by all serious students of this subject without exception?: two studies, for which he cites three references, two of which are newspaper articles (notes 19, 20 and 23 of chapter 6). The first study is by C. R. Snyder, for which Goleman provides two references. Note 19 refers to page 579 of an article by C. Snyder et al. in Journal of Personality and Social Psychology 60, 1991. Note 20 refers to an article that Goleman wrote in the New York Times December 24, 1991, in which he said,
  • Dr. Snyder and his colleagues found that the level of hope among freshman was a more accurate predictor of their college grades than were their S.A.T. scores or their grade point averages in high school The study was reported in part in the November issue of The Journal of Personality and Social Psychology.
  • There is no article by Dr. Snyder in the November issue of this journal in 1990 or 1991. So I can only assume that Goleman means the same article to which he refers in note 19. When I read Goleman's assertions, two objections immediately occurred to me. One is that if hope correlates with high college grades and success in life, then high college grades should predict success in life. The second is that if optimism correlates with freshman college grades, it should also correlate with high school grades, which consequently should be good predictors of college grades. In fact, the only statement concerning this subject on the page that Goleman cites says that high school grades correlate more closely with hope than college grades do: "High school GPA [grade point average] correlated .17, p < .10, and college GPA correlated .13, ns, with Hope Scale scores." On page 582 Snyder et al. report, "Higher Hope Scales correlated -.10, .17, and .49 with better [sic] reported high school performance." The last correlation is higher than any they report with college performance. Nevertheless, the difference among the high school grade correlations is immense, clearly indicating statistical unreliability, which the authors themselves acknowledge (p.582): "one must be cautious in drawing conclusions on the basis of these scant data". Nowhere in the article do Snyder et al. mention SATs.
    The second study Goleman cites is "of five hundred members of the incoming freshman class of 1984 at the University of Pennsylvania, the students' scores on a test of optimism were a better predictor of their actual grades freshman year [sic] than were their SAT scores or their high-school grades." Attacking the massive evidence for the predictive power of SATs and high school grades with a study of five hundred students is like one person attacking an entire army with a butter knife. However, even the evidence to which Goleman refers does not exist. He seriously distorted the reference he cites, which is his own account of this experiment in the New York Times, February 3, 1987, p. C1. There he wrote,
  • Dr. Seligman tested 500 members of the incoming freshman class of 1984. Using a composite of the students' high school grades and college entrance exam scores [i.e. SATs], the dean's office is able to predict what each student's freshman year grades should be. The test of explanatory style, however, was able to predict which freshmen would do better than expected and which would do worse.
  • So this test merely fine-tuned the predictive power of SATs and high school grades among a risibly small sample. Moreover, Goleman's New York Times article provides no reference to an article in a learned journal by which his account can be checked. I see no reason to believe it.
    (Incidentally, the title of the subchapter in which Goleman argues for the importance of optimism is "Pandora's Box and Pollyanna: The Power of Positive Thinking". In it (pp. 86-7) he provides an extremely inaccurate account of what he calls "the familiar legend" of Pandora. It is indicative of how carelessly Goleman reads the sources he cites that the article by Snyder et al. begins with an accurate account of this legend.)

    Goleman's next attack on IQ tests is on page 97, where he extols the importance of empathy. Goleman, typically, provides only one reference for the importance of empathy. That is a not an article or book, but a conference paper. It is not in any university or public library in southern Africa nor in the New York Fifth Avenue Library. Conference papers are usually hard to find. They provide only preliminary reports on research, with little or no supporting evidence. After his reference to this conference paper, Goleman states, without citing any evidence, that children who are good at intuiting other people's feelings "did better in school, even though their IQs were no higher than those of children who are less skilled at reading nonverbal messages - suggesting that mastering this empathic ability smooths the way for classroom effectiveness (or simply makes teachers like them more)." Here again, grades indicate emotional intelligence, contrary to what Goleman says elsewhere. The parenthetical explanation in the sentence quoted is an obvious reason why standardized tests are desirable.

    On page 122, Goleman states, "as tests of children's nonverbal sensitivity have shown, those who misread emotional cues tend to do poorly in school compared to their academic potential as reflected in IQ tests". His footnote refers the reader for this fact to Stephen Nowicki and Marshall Duke's Helping the Child Who Doesn't Fit In. He gives no page numbers and the book has no index. I read it twice and could find no mention of IQ or school performance. It is Goleman's practice never to provide page numbers for books he cites. I have never read another book that cites books without page numbers (although Goleman's hero, Howard Gardner, also does not cite page numbers in the article I discuss above). The lack of page numbers makes references useless for the 99% of readers who do not have the time or interest to look through an entire book for each citation. The explanation for this method of citation that is most favorable to Goleman is laziness and/or stupidity. However, his predilection for citing books without indices seems to indicate deliberate dishonesty, especially since the vast majority of scholarly books, or even popular books on scholarly topics, have indices.

    Chapter 10 is called "Managing with Heart". In its first page (148), Goleman states that an airplane "cockpit is a microcosm of any working organization", so it is a crucially important fact that, "In 80 percent of airline crashes, pilots make mistakes that could have been prevented, particularly if the crew worked together more harmoniously." Typically, Goleman provides no reference to support this assertion, which is antithetical to common sense. On page 149 Goleman reports that, "A study of 250 executives found that most felt their work demanded 'their heads but not their hearts'." Most of these executives spent their entire adult lives in business organizations; all were successful in them. Goleman's proof that they are wrong is a statement he quotes that was made to him and never published (p. 149, note 3).

    Eventually Goleman gets around to denigrating the importance of IQ in running a business or any organization (pp. 160-61):

  • Whenever people come together to collaborate there is a group IQ And how well they accomplish their task will be determined by how high that IQ is. The single most important element in group intelligence, it turns out, is not the average IQ in the academic sense, but rather in terms of emotional intelligence. The key to a high group IQ is social harmony. The idea that there is a group intelligence at all comes from Robert Sternberg and Wendy Williams [They found that] the single most important factor in maximizing the excellence of a group's product was the degree of social harmony.
  • The evidence Goleman cites is an article by Sternberg and Williams on pages 351 to 377 of Intelligence 1988 entitled "Why Some Groups Are Better Than Others". There they do introduce the term "group intelligence", defining it as "the functional intelligence of a group of people working together" (p.356); and they discuss the importance of harmony in one paragraph (p.375). Except for that, what they say has no relation to what Goleman says about group intelligence. That is especially remarkable since Sternberg is, along with Howard Gardner and Stephen Gould, the most prominent academic opponent of the importance of IQ. In fact, Daniel Seligman devotes the third chapter of A Question of Intelligence (1992) to combating Sternberg's attack on IQ. In the article Goleman cites, Sternberg and Williams observe,
  • Groups should contain only the number required by the task, and no more. Groups seldom perform better than their best member would alone (p.352) Past research has uncovered a small relationship between personality measures and measures of group performance (p.353). (Italics added)
  • The groups that Sternberg and Williams studied had three members in each. On page 369 they record the correlations between Group Product Quality and the scores of each member on an IQ test. The correlation for both the highest and second highest scoring members was .65, for the third .43. (The third was clearly being carried by the other two.) These are extremely high correlations, but lower than they would have been in a random sample, since the average IQ of the entire sample was 109, well above the national average (p.365). So the range was narrower than the general population. Sternberg and Williams observe (p.369), "Clearly, IQ is important to doing well on our problems." In the last page, when they sum up, they say, "And finally, IQ [of each member] was an essential component of group intelligence; not only is a lot of IQ on average desirable, but also, one group member particularly high in IQ."
    In addition to completely misrepresenting Sternberg and William's article to make it support his thesis, Goleman also seriously misreports part of it in a way that indicates the cause was simple carelessness or stupidity, not fraudulence. On the bottom of page 160, he says,
  • One surprise was that people who were too eager to take part were a drag on the group, lowering its overall performance; these eager beavers were too controlling or domineering. Such people seemed to lack a basic element of social intelligence, the ability to recognize what is apt and what inappropriate in give-and-take. (Goleman's italics)
  • On page 370 Sternberg and Williams do say, "eager beavers are too controlling and domineering" because they are not "socially apt". But the people they are describing are those who said they would be eager to participate in hypothetical situations like visiting the family of a co-worker who had just died or attending a party at which they know no one and at which the impression they give is crucial to their careers.
    (In the article by Howard Gardner (1995, p. 28) that Goleman cites as proof that IQ accounts for no more than 20% of socioeconomic status Gardner mentions studies by "Sternberg and his colleagues" that he says show that IQ is unimportant in business. He does not name any of these studies.)

    Goleman cites only one other study to support his contention that IQ is unimportant in the functioning of an organization (pp.161):

  • Many things people do at work depend on their ability to call on a loose network of fellow workers. Just how well people can "work" a network is a crucial factor in on-the-job success. Consider, for example, a study of star performers at Bell Labs, the world-famous scientific think tank. The labs are peopled with engineers and scientists who are all at the top on academic IQ tests. But within this pool of talent, some emerge as stars while others are only average in their output. What makes the difference is not their academic IQ, but their emotional IQ. (Goleman's italics)
  • On page 162, Goleman continues to emphasize that the most successful people in this study excelled in networking. This is in line with the importance of harmony he claims was shown in the study by Sternberg and Williams.
    The article he cites ("How Bell Labs Creates Star Performers" by R. Kelley and J. Caplan in pages 128-39 of Harvard Business Review, July-August, 1993) outlines what star performers, who are also called "experts", thought was important for effective job performance and what non-stars, called "middle performers", thought was important. (The designation "middle performers" is a euphemism since there was no third group.) "Taking initiative is the core strategy of the expert model. The second layer of the expert model includes strategies like networking." (pp.132-3). In fact, the experts assigned no more importance to networking than the middle performers did. On page 131 Kelley and Caplan represent what the experts thought is important in the form of a series of concentric circles. In its center is "Core Skills and Strategies: Taking Initiative, Technical Competence, Other Cognitive Abilities" (italics added). These are contrasted with qualities like "Fellowship" in the outer circles.

    Goleman clearly realized that the paramount importance of taking initiative and "other cognitive abilities" conflicts with the crucial role he assigns to harmony and with the title of this chapter ("Managing with Heart"). He mentions taking initiative only once, at the end of the paragraph that concludes his outline of Kelley and Caplan's article:

  • Beyond a mastery of these essential networks, other forms of organizational savvy the Bell stars had mastered included.coordinating their efforts in teamwork building consensus see[ing] things from the perspective of others persuasiveness; and promoting cooperation while avoiding conflict. While all of these rely on social skills, the stars also displayed another kind of knack: taking initiative.
  • So Goleman mentions as if it were a trivial addition what the article he is supposedly outlining emphasizes is the most important skill. He also contrasts that skill with the type of social skills he keeps claiming are vital. However, Kelley and Caplan do not mention any of these examples of "organizational savvy" except persuasiveness. That and the words "organizational savvy" occur on page 133 and are important differences between the two groups. The star performers regarded persuasiveness and "organizational savvy" as the least important abilities, while "middle performers inverted the expert model ranking. According to these engineers, show-and-tell and organizational savvy were the core strategies". This is the opposite of what Goleman claims the article says.
    (The Time review (p. 74) told its readers that this Bell Labs study showed that, "Those workers who were good collaborators and networkers and popular were more likely to reach their goals than socially awkward lone-wolf geniuses.")

    Goleman did, however, quote the article accurately that there was no difference in IQ between the two groups; but he did not quote the reason Kelley and Caplan give (p. 132): "Since all Bell Labs engineers score at the top in IQ tests, cognitive abilities neither guarantee success nor differentiate stars from middle performers." Every champion of the importance of IQ emphasizes that the more important intelligence is for an activity, the narrower will be the range of the IQs of the people doing it and, consequently, the lower the correlation between IQ and successful performance among them. So Arthur Jensen observed that the correlation between IQ and grades decreases from elementary school (.6-.7) to high school (.5-.6) to college (.4-.5) to graduate and professional school (.3). Among colleges, the correlation between SAT scores and grade point average (GPA) is highest at colleges with open admission and lowest at MIT and Caltech. Jensen also points out that there is nearly no correlation between scores on the quantitative part of the Graduate Record Examination and grades of mathematics students at highly selective graduate schools, all of whom have scores in the upper 2% on the quantitative part of the GRE (i.e. in the upper 2% of applicants to graduate schools). (Jensen 1993, pp.151-2; 1980, pp.330-32; 1998, p. 280; 1981, p.30)

    Klitgaard (1985, p.235, n.24 and p.94) uses an analogy from American football. Running speed is the most important ability for a wide receiver. For example, the player personnel director of the New England Patriots said, "We send our five scouts to every school in the country with a draftable player. If he's small or light, we still go if he has the speed. But if he can't run, we don't even look at him." Because professional teams consider only wide receivers who are extremely fast runners, the correlation between their running speed and the order in which they are selected in the professional football draft is only .35. Interestingly, that is similar to the correlation between IQ and grades in graduate and professional schools. There is probably no correlation between speed and performance among the five best professional wide receivers, because they are all extraordinarily fast runners.

    Similarly, as the article Goleman cites points out, the reason that IQ does not correlate with performance at the Bell Labs attests to the paramount importance there of what IQ measures for success. Everyone working in the Bell Labs has an extremely high IQ. In fact, Goleman's introduction of this study, which is quoted above, provides enough information to see that. If any reviewer had taken the effort to read just one book on this subject (including The Bell Curve, pp. 68-9), he would have known that the lack of correlation in this study proves the opposite of what Goleman claims it proves.

    The studies that Goleman cites on pages 190 to 193 to support assertions that there are factors that are more important than IQ are not available in southern Africa or the New York Fifth Avenue Library.

    Goleman's next attack on IQ is on the bottom of page 236 and top of 237: "impulsivity is more directly at cause; impulsivity in ten-year old boys is almost three times as powerful a predictor of their later delinquency as is their IQ". The source Goleman cites for this fact is Jack Block, "On the Relation between IQ, Impulsivity, and Delinquency", Journal of Abnormal Psychology 104 (1995), pp. 395-8. On page 397 of that article, Block quantifies the relative predictive power for delinquency of IQ and impulsivity. He provides four sets of figures: for blacks when impulsivity is entered into the equation before Verbal IQ, for blacks when it is entered after Verbal IQ and for whites when impulsivity is entered before and after Verbal IQ. Only in the first case is impulsivity almost three times as powerful a predictor of their later delinquency as is their IQ. In the second case impulsivity is twice as important. In both cases involving whites, impulsivity and Verbal IQ are about equally correlated with delinquency. (The ratios are 11/10 and 6/5). Since American whites outnumber blacks by eight to one, the ratio for the entire American population is close to the ratio for whites.

    Not only does Goleman misrepresent Block's article, but he also does not tell his readers about the article that immediately follows it (D. Lynam and T. Moffitt, " A Reply to Block (1995)", pp. 399-401). Lynan and Moffitt argue that on the basis of the evidence that Block uses, IQ is much a much better predictor of delinquency than Block concludes it is. They also present evidence against Block's argument, which Goleman repeats on page 335, note 18, that impulsivity causes low IQ. Goleman must have seen Lynam and Moffitt's article and he must have known that anyone who checked his references would see it. If he could have refuted it, he would have done so. He also must have known that anyone with even a slight interest in this topic would be aware of the studies cited in The Bell Curve that show the paramount predictive power of IQ for criminality. These studies are all based on much larger samples than the study Block used. Again the only explanation for Goleman's silence is inability to refute.

    This attack on the importance of IQ is typical of Goleman not only in his gross misrepresentation of the source he cites, but also in using a single, idiosyncratic study and ignoring the massive studies and meta-analyses that are universally accepted by experts but unknown to the general public. For example, the Handbook of Juvenile Delinquency, a standard criminological textbook, edited by H. Quay (1987, pp.106-7), records the results of several meta-analyses:

  • Systematic reviews have concluded that IQ is generally more predictive of offending than social class or cultural background. We know of no current research findings contrary to this conclusion. It has been suggested that official delinquents are more likely to be of lower IQ because they are not clever enough not to be apprehended. Evidence against this position is provided by self-reported delinquency.
  • The last evidence Goleman adduces to denigrate IQ is during his discussion of the crucial importance for young children of being popular with their peers. He says (page 251), "In fact, how popular a child was in third grade has been shown to be a better predictor of mental-health problems at the age of eighteen than anything else - teachers' and nurses' ratings, school performance and IQ, even scores on psychological tests." Again, Goleman totally misrepresented the article he cites to support this statement (E. Cowen et al., "Long-Term Follow-Up of Early Detected Vulnerable Children", Journal of Consulting and Clinical Psychology, 41, 1973, pp.438-46). The study that this article reports has nothing to do with popularity. What it indicates (pp.444-45) is that third grade children's evaluations of their peers' mental problems were the best predictors of future mental problems. Cowen et al. offer the analogy of "recent observation that chronic mental health patients were more sensitive in picking up bogus patients living in hospital wards than mental health professionals". Moreover, they constantly state that previous studies conflict with its conclusions in showing that other factors (teacher and parent judgement, etc.) are accurate predictors of later mental health.
    Goleman cites no later study that supports the conclusions of this article, which was published in 1973. He probably did not bother to search for any. It has no relation to popularity; and even if it did, there would be no reason to try to substantiate it. I have pointed out that anyone who is at all familiar with this subject would immediately realize that even if every study that Goleman cites to show IQ is unimportant was conducted with impeccable rigor and Goleman reported their results accurately, the fact that he could find so few anti-IQ studies and that they all involved minute sample sizes would itself be nearly irrefutable proof that the pro-IQ position is unassailable. Goleman clearly relied on the reviewers of his book being totally ignorant of this subject, not expending the slightest effort to alleviate that ignorance, not reading it carefully enough to see that it is riddled with blatant, irreconcilable contradictions on fundamental issues and not checking a single reference, since if they had checked one reference, they would have found it disturbing enough to check more. Goleman also must have realized that subsequent readers would assume that reviewers did everything they did not do and, consequently, would be predisposed to accept what he says.

    There is one more glaring defect in Goleman's attack on the value of IQ that I have not yet mentioned. He frequently relies on arguments that blatantly misconstrue the nature of statistical correlations and the difference between necessary and sufficient causes. The fact that IQ correlates extremely closely with occupational and personal success does not mean that every person with a high IQ is occupationally and personally successful. The fact that no one with an IQ of 85 can learn high-school-level subjects or be a competent policeman, that no one with an IQ of 100 can be a competent doctor, no one with an IQ of 120 can be a professor of mathematics at MIT and no one with an IQ of 140 can win a Nobel Prize in Physics does not mean that a person with an IQ of 200 will do any of these.

    To be more specific, in a study of a representative group of policemen and fireman the lowest IQ was 96, and most studies of white medical students and doctors find none with IQs below 115. These statistics can be amplified by a study done by the United States Department of Labor in which a representative sample of 39,600 employed people were given the US Employment Services General Aptitude Test Battery (GATB). Scores on this test are highly correlated with scores on IQ tests, but it magnifies differences. A score of 140 on the GATB equals a score of 130 on the Wechsler IQ tests, and 60 on the GATB equals an IQ of 70. The average scores of the members of different occupations ranged from 55 for potato peelers to 143 for mathematicians. But the difference between the lowest score of members of low-IQ occupations and the lowest score of members of high-IQ occupations was 86 points, while the difference between the highest scores was only 12. (Jensen 1980, pp. 341-4; Gordon 1988, p. 90n9) Hans Eysenck (1998, p. 21) provides a chart with the average IQ and the standard deviation (SD) in IQ of members of various occupations. The higher the average IQ is, the smaller the SD is. The highest IQ occupations on his chart, which is not inclusive, are lawyer, accountant and auditor, which have average IQs of 127-8 and SDs of 10.9-11.7. The nine occupations he lists whose members have an average IQ below 100, have SDs between 18.7 and 20.8.

    People with high IQs can peel potatoes, but people with low, or even normal, IQs cannot be mathematicians or lawyers. However, the average scores for the low-IQ occupations show that extremely few high IQ people are doing them; and most high-IQ people who do work at them, do so temporarily, while they are students, or between jobs.

    Goleman uses these facts to make statements that no one disputes but which seem to support his case; for example (p.34): "The brightest among us can founder people with high IQs can be stunningly poor pilots of their private lives. Many people with very low IQs end up in menial jobs, and those with high IQs tend to become well paid - but by no means always."

    At other times he makes statements that are demonstrably wrong but which the reviewers quoted as if they had been proved; for example, on page 41 he quotes Howard Gardner with approval: "many people with IQs of 160 work for people with IQs of 100". (Typically, Goleman supplies no reference to where Gardner said this.) There are people with IQs of 160 doing low-level work and people with IQs of 100 who own businesses. But in both cases the number is negligible. The businesses that are owned by people with IQs of 100 must be either extremely simple, like fruit stalls, or businesses they inherited and allow other people to run. So there may be one or two people in the United States with IQs of 160 who work for people with IQs of 100, but not "many". The average IQ of high school graduates is 105-106. That means that a person with a 100 IQ has to struggle to finish high school; but only 7% of people with only high school diplomas have as high an IQ as the average college graduate and only 1% as high as the average person with a PhD, MD, or LLB. (Those people with IQs of 85 and below who have obtained high school diplomas score no higher than fifth or sixth grade level on standard achievement tests.) (Herrnstein and Murray.1994, pp.49, 151-2; Jensen 1998, p. 553. No one disputes any of these statistics.)

    Before ending my analysis of Goleman's attacks on IQ, I will return once again to the crucial Marshmallow Test. I have shown that Goleman grossly distorts the other studies that he claims denigrate the importance of IQ and academic performance. The Marshmallow Test is not an exception. Goleman outlines it in a chapter entitled "The Master Aptitude" (self-control) and a section entitled "Impulse Control: The Marshmallow Test", in which he says that the marshmallow test shows that, "There is perhaps no psychological skill more fundamental than resisting impulse. It is the root of all emotional self-control" (p.81).

    However, the article Goleman cites for the Marshmallow Test (Y. Shoda, et al., "Predicting Adolescent Cognitive and Self Regulatory Competencies From Preschool Delay of Gratification ", Developmental Psychology 26, 1990, pages 978-86) reports that the results Goleman outlines were not for self-control. According to this article, Goleman's source, the children were exposed to various desirable objects, only one of which were marshmallows. They were divided into four groups, distinguished by whether the rewards were exposed or hidden and whether the children were advised as to how to distract themselves or were not. The remarkable correlations between adolescent emotional, social and academic strengths and SAT scores that Goleman emphasizes did exist, but only for those children who were exposed to the rewards and were not advised on how to cope with the temptation. For the other three groups, there was no significant correlation among these adolescent traits, and the correlations between them and childhood resistance to temptation were mostly negative. So the superior adolescent social, emotional and academic performance that Goleman says is so important were not predicted by children's ability to resist temptation, but by their ability to devise cognitive strategies. The last paragraph of the article Goleman cites concludes, "cognitive and attentional strategies and skills play an important role in the delay situation used in the present study". However, it then refers to other studies that show that "there is also much evidence that other factors are likewise germane for a comprehensive analysis of delay of gratification". Moreover, as with nearly every other article Goleman cites, this one warns on its last page, "We must emphasize the need for caution in the interpretation of the total findings given the smallness of the sample".

    Goleman's arguments for devoting even more of American education to emotional and social skills than at present are as patently absurd as his attacks on the importance of IQ, school grades and achievement tests. His main argument is on pages 231 to 233 and the footnotes to them (p. 333, notes 3 to 5; the italics are his.):

  • In 1990, compared to the previous two decades, the United States saw the highest juvenile arrest rate for violent crimes ever; teen arrests for forcible rape had doubled; teen murder rates quadrupled. During the same two decades, the suicide rate for teenagers tripled. As of 1993 the birthrate among girls ten to fourteen has risen steadily for five years in a row. Heroin and cocaine use among white youth climbed about 300 percent over the two decades before the 1990s; for African-American youth it jumped to a staggering 13 times the rate of twenty years before. Symptoms of depression affect up to one third of teenagers. The frequency of eating disorders in teenage girls has skyrocketed. In a national sample of American children, ages seven to sixteen, comparing their emotional condition in the mid-1970s and at the end of the 1980s there was a steady worsening [of] all [emotional] indicators (pp.231-3). Teen arrest rates for forcible rape rose from 10.9 per 100,000 in 1965 to 21.9 in 1990. Teen murder rates more than quadrupled from 1965 to 1990. In 1950 the suicide rate for those 15 to 24 was 4.5 per 100,000. By 1989 it was three times higher. Over the three decades since 1960 rates of gonorrhea jumped to a level four times higher among children 10 to 14, and three times higher among those 15 to 19.
  • These and similar observations about the recent collapse of adolescent emotional health on pages 240 to 241 are in the penultimate chapter of Emotional Intelligence, entitled "The Cost of Emotional Illiteracy". It and the last chapter, "Schooling the Emotions", where Goleman explains his educational proposals, comprise the last section, "Emotional Literacy". Similarly in the Preface, Goleman says (pp. xiii-xlv),
  • perhaps the most disturbing single piece of data in this book [is that] the present generation of children [is] more troubled emotionally than the last. One solution is education [that] include[s] essential competencies such as self-awareness, self-control, and empathy, and the arts of listening, resolving conflicts, and cooperation.
  • Yet, clearly these statistics on the emotional collapse of American adolescents in the past several decades are extremely powerful evidence that Goleman's position is not just wrong but nonsensical. Expenditure on American education, per student and adjusted for inflation, rose 58% in the 1960s, 27% in the 1970s, 29% in the 1980s (the "decade of greed"). Between 1960 and 1990 the average American class size decreased by a third; enrolment declined by 7%, but the number of teachers rose by 17%. Then, between the academic year 1989-1990 and 1994-95, the expenditures of all educational institutions in the USA increased from 381 billion dollars to 508 billion dollars. (Sowell 1993, pp.12, 270; D'Souza 1995, p.649n3; Sowell 1993, pp.3-4,7-9 and passim; US Digest of Educational Statistics)
    In return for this vast sum of money, the American people have not received improved academic performance (D'Souza 1995, p.649n3). I pointed out above that in 1981 the average SAT score of New York State students was seven points above the national average; in 1995 it was nineteen points below. In 1995 NY State spent 56% more than the national average on education per student. Between 1981 and 1995 its expenditure per student per year increased from less than $4000 to $9300. Even more striking is that beginning in 1987, under court order, the state of Kansas subsidized 56 Kansas City public schools an average of $36,111 per student per year above what they had been spending. The result was that in these schools between 1987 and 1992 the proportion of students graduating dropped from 58% to 38%, and scores on standardized mathematics and reading tests plummeted. (Economist June 22, 1996, p.58; August 28, 1993)

    A major reason why expenditure brings no return is the tremendous amount of time, energy and money that is wasted on the futile attempt to teach elementary-school-level subjects to people with IQs of 75, high-school-level subjects to people with IQs of 85 and college-level-subjects to people with IQs of 100, along with the tremendous amount of time, energy and money that is wasted pretending that they are learning something. (In his 1997 State of the Union Speech, President Clinton proclaimed the goal that "every eighteen-year-old must be able to go to college".)

    Another reason why the phenomenal increase in expenditure on education has produced no improvement in academic performance since 1960 must be the proportion of education devoted to non-academic activities. I mention above that in 1994 less than one-third of the education budget of NY City was for classroom education. A great deal of the rest was for psychologists, guidance councillors, etc., many of whom are specialists in emotional education. Also in 1994, only 41% of the average American school day was spent on academic subjects and the average American high school student spent much less than half the number of hours studying academic subjects than the average French or Japanese high school student spent and less than a third of the German average. Outside of school, the average American high school student spends four hours a week doing homework, as opposed to an average of four hours a day in other economically advanced countries. Yet adolescents in other countries have much lower rates of emotional problems than American adolescents have. (Sykes 1995, pp.16, 228; Thernstrom 1997, p. 382; Economist, June 22, 1996, p.58)

    The concentration of American education on non-academic concerns is the result of a process that has been in progress for many decades despite opposition from parents, who have sometimes been supported by teachers and the government officials they elect. It has been led by the National Education Association (NEA), which is a professional association, trade union and "the only union that owns its own cabinet department". Its immense power has been exercised despite the fact that most teachers do not agree with it on key issues. Among the NEA's positions is that standardized tests are "similar to narcotics" in "maiming" children. In 1918 a commission formed by the NEA issued a statement, which is still quoted, that schools should concern themselves more with "preparation for effective living" than with academics. However, it was not until the end of World War II that the US Department of Education adopted the ideal of "life adjustment" and "personal satisfaction". These ideals transformed American schools in the late 1940s and early 1950s. In 1947 the extremely influential yearbook of the NEA's Association for Supervision and Curriculum Development, now representing orthodox opinion, stated, "Far too many people in America look upon the elementary school as a place to learn reading, writing and arithmetic [Schools must put] human relationships first"; "It is the responsibility of the schools to be alert to the symptoms of strong emotions". Consequently, "We are going to have to change our ideas about the things we expect from teachers She will help the children learn how to work together She will listen to each child and will help find what he needs to grow". The new approach required not only a reorientation of what teachers do, but also an army of non-teachers to minister to children's emotional and social needs and problems. Between 1960 and 1991 the proportion of the staff of American schools who were not teachers rose from 25% to 47%. (Sykes 1995, p.197-201, 205-7, 215, 228-33)

    Not only were many non-academic subjects introduced after World War II, but school texts were also radically simplified between the late 1940s and early 1960s. The words they used were simplified, and the average sentence-length in sixth, seventh and eighth grade readers declined from twenty words before World War II to fourteen in 1993. (Zajonc and Mullally 1997, p.696; Sykes 1995, pp.128-30)

    The 1947 NEA yearbook promised that the reward for following their approach will be, "Poverty, malnutrition, economic injustice, intolerance, ignorance will all yield to a dynamic program of education in the hands of socially literate teachers." (Sykes 1995, pp.197-8) As Goleman shows so convincingly, it is exactly when American schools were converted to teaching emotional and social skills that the emotional and social health of American adolescents began their precipitous decline. Of course, the NEA and other powerful forces and people in the educational establishment have continued to champion more emotional and educational education even though statistics like the ones that Goleman provides on the emotional collapse of American adolescents prove conclusively that such education is ineffectual (Sykes 1995, pp.33, 45-7, 229, 246-9).

    It is difficult for most people today to comprehend how thoroughly American education has been re-routed away from academics. The best illustration is provided by what used to be required of American students.

  • In the early years of the twentieth century, pupils finishing the eighth grade in Kansas had to pass an examination which included spelling such words as "elucidation" and "animosity", defining such terms as "zenith" and "panegyric" and doing such problems in arithmetic as finding the interest earned on a $900 note, at 8 percent, after 2 years, 2 months, and 6 days. Questions of similar difficulty were asked in geography and history - all in order to get a diploma at the end of the eighth grade. Often [these schools] were one-room school houses. (Sowell 1993, pp. 7-8; italics added)
  • Similarly, the following are typical questions on an examination required for admission to Jersey City High School in 1885 (Sykes 1995, p.61; Herrnstein and Murray 1994, p.419):
  • Define Algebra, an algebraic expression, a polynomial. Make a literal trinomial.
  • Write a homogeneous quadrinomial of the third degree
    Find the sum and difference of 3x-4y+7cd-4xy+16 and 10ay-3x-8xy+7cd-13.

    Find the product of 3+4x+5x2-6x3 and 4-5x-6 2

    Write a sentence containing a noun used as an attribute, a verb in the perfect tense potential mood, and a proper adjective.

    Name the four principle ranges of mountains in Asia, three in Europe, and three in Africa.

    Name the capitals of the following countries: Portugal, Greece, Egypt, Persia, Japan, China, Canada, Tibet and Cuba. [It is interesting that these questions are in no way American-Euro-centric.]

    Name three events of 1777. Which was the most important and why?

    As for the courses that were taken in high school, in Thornton Wilder's Our Town (beginning of Act II) it is assumed that in 1904 every student in a high school in a small New Hampshire town, nearly none of whom would attend college, studied solid geometry and Cicero's Orations. By contrast, in 1983 less than a third of high school graduates had taken a course in intermediate algebra (Sykes 1995, p.238).
    A much smaller proportion of adolescents were in school then than now, but a much higher proportion of adolescents of that age could answer the above questions then than now. The students who answered these questions were taught by teachers who had never taken an education course and never considered trying to teach emotional or social skills (Toch 1991, pp.42-3, 46-51; Stevenson and Stigler 1992, pp.107-8).

    The efficacy of education courses is not a matter for speculation or debate. Since the 1970s, many American states have required prospective public school teachers to pass standardized qualifying examinations. The core of these tests do not examine specific job knowledge or performance, but reading comprehension, writing ability, solving mathematical problems, identifying main ideas and sequential steps, drawing inferences, etc. In other words, they are intelligence tests, and teachers' scores on them correlate with their SAT scores. Numerous careful studies have shown no positive correlation between student performance and educational expenditure, student motivation, post-high-school educational intention, self-esteem or any other factor except one. That is an extremely robust correlation between student performance and their teachers' scores on these examinations.

    The most extensive study, of 105 school districts in North Carolina, found that among students of all races and socioeconomic backgrounds, a difference between school districts of only one percent in their teachers' average score on the National Teacher Evaluation examination produced a five per cent lower failure rate of high school juniors on standardized reading and mathematics tests. No other factor had any effect. (Strauss and Sawyer 1986)

    Similarly, a difference of one standard deviation (SD) on the Texas Examination of Current Administrators and Teachers (TECAT) causes a .17 SD difference in student scores on standardized academic achievement tests over a period of two years. That in itself is significant. But the effect is cumulative. The longer students are taught by teachers with higher or lower TECAT scores, the greater the effect it has on them. As on other teacher qualifying tests, a teacher's score on the TECAT correlates with his SAT score. (Ferguson 1991, pp. 471-6, 482; Ferguson 1998, pp. 350-57)

    So in teaching, as in every other occupation, success is correlated with a score on a test of general intellectual ability, not of specific job performance, let alone attitude, enthusiasm, emotional health or social skills.

    The facts above are well known and have been much discussed. There can be only two reasons for Goleman ignoring them: either he is totally ignorant of even the most elementary facts and controversies in the subject on which he is pontificating; or he could not refute their obvious implications and knew that he did not have to because his book would be reviewed by complete ignoramuses and that later readers would assume that the reviewers' evaluations were based on knowledge of this subject.

    I will now add a heterodox opinion. The people who publicize the deficiencies of American academic education, especially compared with that of other countries, assume that this inferiority will eventually cause economic decline, and maybe even collapse. That is a logical expectation, but it is contradicted by empirical evidence. The Economist of March 29, 1997 (pages 21 to 25) reported the results on tests of mathematical and scientific knowledge and ability taken by 500,000 thirteen-year-olds around the world. It said,

  • President Clinton described the test [sic] in his state-of-the-union message as one "that reflects world-class standards our children must meet" America's poor performance sparked calls for the adoption of a national curriculum and national standards - including from Mr. Clinton himself.
  • The Economist then reported similar reactions in France and Germany, whose children also did poorly, and it assumed that the people of these countries and their leaders are right to be worried.
    However, the statistics the Economist presents show no correlation between performance on these tests and economic growth. Six of the fifteen best-performing countries in both math and science are ex-Communist countries. (The Czech Republic was second in the world in science and sixth in math.) Everyone who lived in the United States in the years following the launch of the Sputnik in 1957 remembers the horrible anguish, even terror, caused by the superiority of Soviet over American education. Soviet education remained much better than American. The mean score of Russians on the Physics Graduate Record Exam between 1992 and 1995 was over 750, compared to approximately 600 for Americans (Glanz 1996, p. 710). But by then it was obvious that superior Russian education was economically irrelevant. In the 1980s, as the Japanese economic hare seemed to streak past the American economic tortoise, the superiority of Japanese education also caused considerable worry, and Japanese thirteen-year-olds were third in both math and science on the 1997 tests. Now it seems that Japanese educational superiority is also irrelevant. Despite mediocre American education, the American economy is the strongest in the world, and it is especially strong in high technology. As long as American capital markets are efficient, scientists and engineers can be imported. In 1996 58% of doctoral recipients from American universities in Engineering, 47% in Physical Sciences and 38% in Life Sciences were non-Americans. More than 70% of foreign doctorate recipients remain in the United States. Critics of American society frequently observe that investment bankers and mergers-and-acquisition experts make many times more money than scientists and engineers. However, investment bankers and mergers-and-acquisitions experts are much more valuable economically than scientists and engineers; and they require very little formal education. (Henderson et al. 1998, p. 17; Bhawgwati 1994)

    The last page of Emotional Intelligence, entitled "About the Author", states, "Daniel Goleman, Ph.D., covers the behavioral and brain sciences for The New York Times and his articles appear throughout the world in syndication." Of all the statements in this book that I could check, this is the only one that I found to be correct. The New York Times is the most read newspaper among the people who shape American society (members of Congress, corporate executives, heads of major labor unions, etc.) and, even more importantly, among the producers and writers of television news programs, which are the main source of news for most Americans (S. Lichter, et al. 1986, pp. 11-12). It also exerts a tremendous influence on the major news organizations in deciding which issues merit attention (Lynch 1989, pp.96-7, 108n3). This is the quality of the media's coverage of these topics, which are manifestly of crucial importance.
     
     
     
     
     
     

    BIBLIOGRAPHY

    C. P. Benbow 1988: "Sex Differences in Mathematical Reasoning Ability in Intellectually Talented Preadolescents", The Behavioral and Brain Sciences, 11,2, pp.169-232.

    C. P. Benbow 1992: "Academic Achievement in Mathematics and Science of Students between Ages 13 and 23 ", Journal of Educational Psychology 84, pp.51-61.

    J. Bhagwati 1994: "Foreign Students Spur U.S. Brain Gain", Wall Street Journal Europe, June 10.

    C. Brand 1996: "Doing Something about g", Intelligence 22, pp.311-26.

    N. Brody 1992: Intelligence (second edition).

    J. Carroll 1997: "Psychometrics, Intelligence and Public Perception", Intelligence 24, pp. 25-52.

    S. Coren 1994: The Intelligence of Dogs.

    T. Eastland 1996: Ending Affirmative Action.

    H. J. Eysenck 1998: A New Look at Intelligence, Transaction Publishers, New Brunswick, New Jersey.

    J. Fallows 1982: "The Tests and the 'Brightest': How Fair are the College Boards", Atlantic, February, pp.38-49.

    D. Farber and S. Sherry 1997: Beyond All Reason.

    R. F. Ferguson 1991: "Paying For Public Education", Harvard Journal on Legislation 28, pp.465-98.

    R. F. Ferguson 1998: "Can Schools Narrow the Black-White Test Score Gap?", in C. Jencks and M. Phillips (editors), The Black-White Test Score Gap, The Brookings Institute, Washington D. C., pp. 318-74.

    M. Frydman and R. Lynn 1989: "The Intelligence of Korean Children Adopted in Belgium", Personality and Individual Differences 10, pp.1323-25).

    H. Gardner, "Cracking open the IQ Box" in pp. 23-35 of The Bell Curve Wars, edited by S. Fraser.

    J. Glanz 1996: "How not to Choose a Physicist?", Science, 274, 1 November.

    R. Gordon 1984: "Digits Backward and The Mercer-Kamin Law" in C. Reynolds and R. Brown (editors), Perspectives on Bias in Mental Testing, pp.357-506.

    R. Gordon 1988: "Thunder from the Left", Academic Questions 1, pp. 75-92.

    E. Hanushek 1991: "When School Finance 'Reform' May Not Be Good Policy", Harvard Journal on Legislation 28, pp. 423-56.

    J. Heckman 1995: "Lessons from the Bell Curve", Journal of Political Economy 103,5, pp.1091-1120.

    P. H. Henderson, J. E. Clarke, C. Woods 1998: Summary Report 1996: Doctoral Recipients from United States Universities.

    R. Herrnstein and C. Murray 1994: The Bell Curve. The addition of an Afterword (sic) on pages 553-75 of the 1996 edition makes its subsequent pagination different from the original, 1994, edition.

    A. Jensen 1980: Bias in Mental Testing.

    A. Jensen 1981: Straight Talk about Mental Tests.

    A. Jensen 1985: "The Nature of the Black-White Difference on Various Psychometric Tests: Spearman's Hypothesis", The Behavioral and Brain Sciences 8, pp. 193-263.

    A. Jensen 1993: "Psychometric G and Achievement" in B. Gifford (editor), Policy Perspectives on Educational Testing.

    A. Jensen 1998: The g-Factor.

    R. Klitgaard 1985: Choosing Elites.

    P. Legree 1995: "Evidence for an Oblique Intelligence Factor" Intelligence 21, pp.247-66.

    M. Levin 1997: Why Race Matters.

    S. Lichter, S. Rothman, and L. Lichter 1986: The Media Elite.

    R. Lynch 1989: Invisible Victims: White Males and the Crisis of Affirmative Action.

    R. Lynn 1993: "Oriental Americans: Their Educational Attainment and Socio-Economic Status", Personality and Individual Differences 15, pp. 237-42.

    R. Lynn 1996: "Racial and ethnic Differences in Intelligence in the United States on the Differential Ability Scale", Personality and Individual Differences 20, pp. 271-3.

    R. Lynn 1997: "Intelligence in Taiwan", Personality and Individual Differences 22, pp.585-6.

    R. Lynn and M. Song 1994: "General Intelligence, Visuospatial and Verbal Abilities in Korean Children", Personality and Individual Differences 16, pp.363-4.

    W. Manning and R. Jackson 1984: "College Entrance Examinations" in C. Reynolds and R. Brown (editors), Perspectives on Bias in Mental Testing.

    M. Mayer 1967: The Lawyers.

    J. Mayer and G. Geher 1996: "Emotional Intelligence and the Identification of Emotion" Intelligence 22, pp.89-113.

  • C. Murray 1995: "IQ, Race, and Heredity", Commentary August, pp.15-25.
  • C. Murray 1995A: "'The Bell Curve' and its Critics", Commentary, May, pp. 23-30.
    P. Mussen, S. Harris, S. Rutherford and C. Keaseey 1970: "Honesty and Altruism among Preadolescents", Developmental Psychology 3, pp.167-91.

    C. Murray 1996: "Afterword" to the 1996 paperback edition of The Bell Curve, pp.553-75.

    C. Murray 1997: "IQ and Economic Success", The Public Interest, 128, Summer, pp. 21-35.

    C. Murray and R. Herrnstein 1994: "Race, Genes and IQ - An Apologia" The New Republic October 31, pp. 27-37.

    U. Neisser et al 1996: "Intelligence: Knowns and Unknowns", American Psychologist, 54, pp. 77-101.

    H. Quay (editor) 1987: Handbook of Juvenile Delinquency.

    D. Oren 1985: Joining the Club: A History of Jews of Jews and Yale.

  • J. Philippe Rushton 1995: Race, Evaluation and Behavior.
  • Y. Shoda, W. Mishel and P. Peake 1990: "Predicting Adolescent Cognitive and Self-Regulatory Competencies from Preschool Delay of Gratification", Developmental Psychology 26, pp.978-86.
    D. Seligman 1992: A Question of Intelligence.

    W. Sloshberg and W. NesSmith 1983: Contemporary American Society.

    M. Snyderman and S. Rothman 1988: The IQ Controversy: the Media and Public Policy.

    T. Sowell 1993: Inside American Education.

    S. Steinberg 1974: The Academic Melting Pot.

    S. Steinberg 1981: The Ethnic Myth.

    H. Stevenson and J. Stigler 1992: The Learning Gap: Why Our Schools Are Failing and What We Can Learn from Japanese and Chinese Education.

    M. Storfer 1990: Intelligence and Giftedfness.

    R. Strauss and E. Sawyer 1986: "Some New Evidence on Teacher and Student Competencies", Economics of Education Review 5, pp. 41-8.

    C. Sykes 1995: Dumbing Down Our Kids: Why American Children Feel Good About Themselves But Can't Read, Write, or Add.

    M. Synnott 1979: The Half-Opened Door.

    S. and A. Thernstrom 1997: America in black and White.

    S. Thernstrom 1998A: "Diversity and Meritocracy in Legal Education", Constitutional Commentary, 15,1, pp. 11-43.

    S. Toch 1991: In the Name of Excellence

    G. Vaillant 1977: Adaptation to Life.

    P. Vernon 1982: The Abilities and Achievements of Orientals in North America.

    N. Weyl 1989: The Geography of American Achievement.

    J. Wilson and R. Herrnstein 1985: Crime and Human Nature.

    R. Zajonc and P. Mullallay 1997: "Birth Order, Reconciling Conflicting Effects", American Psychologist 52, pp. 685-99.

    Torna alla pagina principale