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
(Charles Darwin, The Descent of Man)
No "serious" book published in the 1990s received higher
praise from the press than Daniel Goleman's
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
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,
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
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,
Below are excerpts from some reviews of Emotional
Intelligence. Thefirst is from the New York Times,
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 courseBut
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
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
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
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
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
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
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.
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
"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
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.
"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
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".
"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
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
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
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
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.
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.
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
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
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
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
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
However, Time was talking not about success at
performing a job, but success as measured by the status of one's job. (Time
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
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
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
Bell Curve, which never makes a statement on this subject, or any other,
without discussing the extremelyextensive 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
The third and fourth statements so blatantly and totally distort what
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
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
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 everykind 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 higheraverage 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
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
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
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
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'sis 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,
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 Universityin 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;
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,
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
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
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,
Educational Review 39, 1969, p. 525)
However, at least The U.S. News and World Report
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".
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,
(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
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
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
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,
< .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
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
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",
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
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 IQthat I have not yet mentioned.
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
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
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
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 $36,111per 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 diplomaat the end of the eighth grade.
Often [these schools] were one-room school houses. (Sowell 1993, pp. 7-8;
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
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
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
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 1957remembers
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
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
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",
Street Journal Europe, June 10.
C. Brand 1996: "Doing Something about g",
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
H. Gardner, "Cracking open the IQ Box" in pp. 23-35 of
Bell Curve Wars, edited by S. Fraser.
J. Glanz 1996: "How not to Choose a Physicist?", Science,
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",
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
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
M. Mayer 1967: The Lawyers.
J. Mayer and G. Geher 1996: "Emotional Intelligence and
the Identification of Emotion" Intelligence
C.Murray 1995: "IQ, Race, and Heredity", Commentary August,
C. Murray 1995A: "'The Bell Curve' and its Critics",
May, pp. 23-30.
P. Mussen, S. Harris, S. Rutherford and C. Keaseey
1970: "Honesty and Altruism among Preadolescents", Developmental Psychology
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
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
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
R. Zajonc and P. Mullallay 1997: "Birth Order, Reconciling
Conflicting Effects", American Psychologist 52, pp. 685-99.