I.

Introduction

Print section

Intelligence, term usually referring to a general mental capability to reason, solve

problems, think abstractly, learn and understand new material, and profit from past
experience. Intelligence can be measured by many different kinds of tasks. Likewise, this
ability is expressed in many aspects of a person's life. Intelligence draws on a variety of
mental processes, including memory, learning, perception, decision-making, thinking, and
reasoning.

II. Defining Intelligence

Print section

Most people have an intuitive notion of what intelligence is, and many words in the English
language distinguish between different levels of intellectual skill: bright, dull, smart, stupid,
clever, slow, and so on. Yet no universally accepted definition of intelligence exists, and
people continue to debate what, exactly, it is. Fundamental questions remain: Is intelligence
one general ability or several independent systems of abilities? Is intelligence a property of
the brain, a characteristic of behavior, or a set of knowledge and skills?

The simplest definition proposed is that intelligence is whatever intelligence tests measure.
But this definition does not characterize the ability well, and it has several problems. First, it
is circular: The tests are assumed to verify the existence of intelligence, which in turn is
measurable by the tests. Second, many different intelligence tests exist, and they do not all
measure the same thing. In fact, the makers of the first intelligence tests did not begin with
a precise idea of what they wanted to measure. Finally, the definition says very little about
the specific nature of intelligence.

Whenever scientists are asked to define intelligence in terms of what causes it or what it
actually is, almost every scientist comes up with a different definition. For example, in 1921
an academic journal asked 14 prominent psychologists and educators to define intelligence.
The journal received 14 different definitions, although many experts emphasized the ability
to learn from experience and the ability to adapt to one's environment. In 1986 researchers
repeated the experiment by asking 25 experts for their definition of intelligence. The
researchers received many different definitions: general adaptability to new problems in
life; ability to engage in abstract thinking; adjustment to the environment; capacity for
knowledge and knowledge possessed; general capacity for independence, originality, and
productiveness in thinking; capacity to acquire capacity; apprehension of relevant
relationships; ability to judge, to understand, and to reason; deduction of relationships; and
innate, general cognitive ability.

People in the general population have somewhat different conceptions of intelligence than
do most experts. Laypersons and the popular press tend to emphasize cleverness, common
sense, practical problem solving ability, verbal ability, and interest in learning. In addition,
many people think social competence is an important component of intelligence.

Most intelligence researchers define intelligence as what is measured by intelligence tests,

but some scholars argue that this definition is inadequate and that intelligence is whatever
abilities are valued by one's culture. According to this perspective, conceptions of
intelligence vary from culture to culture. For example, North Americans often associate
verbal and mathematical skills with intelligence, but some seafaring cultures in the islands
of the South Pacific view spatial memory and navigational skills as markers of intelligence.
Those who believe intelligence is culturally relative dispute the idea that any one test could
fairly measure intelligence across different cultures. Others, however, view intelligence as a
basic cognitive ability independent of culture.

In recent years, a number of theorists have argued that standard intelligence tests measure
only a portion of the human abilities that could be considered aspects of intelligence. Other
scholars believe that such tests accurately measure intelligence and that the lack of
agreement on a definition of intelligence does not invalidate its measurement. In their view,
intelligence is much like many scientific concepts that are accurately measured well before
scientists understand what the measurement actually means. Gravity, temperature, and
radiation are all examples of concepts that were measured before they were understood.

III. Measuring Intelligence

Print section

The first intelligence tests were short-answer exams designed to predict which students
might need special attention to succeed in school. Because intelligence tests were used to
make important decisions about people's lives, it was almost inevitable that they would
become controversial. Today, intelligence tests are widely used in education, business,
government, and the military. However, psychologists continue to debate what the tests
actually measure and how test results should be used.

A. Early Tests

Print section

Interest in measuring individual differences in mental ability began in the late 19th century.
Sir Frances Galton, a British scientist, was among the first to investigate these differences.
In his book Hereditary Genius (1869), he compared the accomplishments of people from
different generations of prominent English families. No formal measures of intelligence
existed at the time, so Galton evaluated each of his subjects on their fame as judged by
encyclopedia entries, honors, awards, and similar indicators. He concluded that eminence of
the kind he measured ran in families and so had a hereditary component. Believing that
some human abilities derived from hereditary factors, Galton founded the eugenics
movement, which sought to improve the human species through selective breeding of gifted
individuals.

Between 1884 and 1890 Galton operated a laboratory at the South Kensington Museum in
London (now the Victoria and Albert Museum) where, for a small fee, people could have
themselves measured on a number of physical and psychological attributes. He tried to
relate intellectual ability to skills such as reaction time, sensitivity to physical stimuli, and
body proportions. For example, he measured the highest and lowest pitch a person could
hear and how well a person could detect minute differences between weights, colors,
smells, and other physical stimuli. Despite the crude nature of his measurements, Galton
was a pioneer in the study of individual differences. His work helped develop statistical
concepts and techniques still in use today. He also was the first to advance the idea that
intelligence can be quantitatively measured.

In the 1890s American psychologist James McKeen Cattell, who worked with Galton in
England, developed a battery of 50 tests that attempted to measure basic mental ability.
Like Galton, Cattell focused on measurements of sensory discrimination and reaction times.
Cattell's work—and by association, Galton's—was unsupported in 1901, when a study
showed that the measurements had no correlation with academic achievement in college.
Later researchers, however, pointed out that Cattell's test subjects were limited to Columbia
University students, whose high academic performance was not representative of the
general population. Better-designed tests given to broader samples have shown that
reaction time and processing speed on perceptual tasks do correlate with academic
achievement.

B. The Binet-Simon Test

Print section

Alfred Binet, a prominent French psychologist, was the first to develop an intelligence test
that accurately predicted academic success. In the late 19th century, the French
government began compulsory education for all children. Prior to this time, most
schoolchildren came from upper-class families. With the onset of mass education, French
teachers had to educate a much more diverse group of children, some of whom appeared
mentally retarded or incapable of benefiting from education. Teachers had no way of
knowing which of the “slow” students had true learning problems and which simply had
behavioral problems or poor prior education. In 1904 the French Ministry of Public
Instruction asked Binet and others to develop a method to objectively identify children who
would have difficulty with formal education. Objectivity was important so that conclusions
about a child's potential for learning would not be influenced by any biases of the examiner.
The government hoped that identifying children with learning problems would allow them to
be placed in special remedial classes in which they could profit from schooling. Binet and
colleague Théodore Simon took on the job of developing a test to assess each child's
intelligence.

As Binet and Simon developed their test, they found that tests of practical knowledge,
memory, reasoning, vocabulary, and problem solving worked better at predicting school
success than the kind of simple sensory tests that Galton and Cattell had used. Children
were asked, among other tasks, to perform simple commands and gestures, repeat spoken
digits, name objects in pictures, define common words, tell how two objects are different,
and define abstract terms. Similar items are used in today's intelligence tests. Binet and
Simon published their first test in 1905. Revisions to this test followed in 1908 and 1911.

Binet and Simon assumed that all children follow the same course of intellectual
development but develop at different rates. In developing their test, they noted which items
were successfully completed by half of seven-year-olds, which items by half of eight-year-
olds, and so on. Through these observations they created the concept of mental age. If a
10-year-old child succeeded on the items appropriate for 10-year-olds but could not pass
the questions appropriate for 11-year-olds, that child was said to have a mental age of 10.
Mental age did not necessarily correspond with chronological age. For example, if a 6-year-
old child succeeded on the items intended for 9-year-olds, then that child was said to have a
mental age of 9.
To judge how effectively the test predicted academic achievement, Binet asked teachers to
rate their students from best to worst. The results showed that students who had been
rated higher by their teachers also scored higher on the test. Thus, Binet's test successfully
predicted how students would perform in school.

C. The IQ Test

Print section

Binet's test was never widely used in France. Henry Goddard, director of a New Jersey
school for children with mental retardation, brought it to the United States. Goddard
translated the test into English and began using it to test people for mental retardation.
Another American psychologist, Lewis Terman, revised the test by adapting some of Binet's
questions, adding questions appropriate for adults, and establishing new standards for
average performance at each age. Terman's first adaptation, published in 1916, was called
the Stanford-Binet Intelligence Scale. The name of the test derived from Terman's affiliation
with Stanford University.

Instead of giving a person's performance on the Stanford-Binet as a mental age, Terman

converted performance into a single score, which he called the intelligence quotient, or IQ.
A quotient is the number that results from dividing one number by another. The idea of an
intelligence quotient was first suggested by German psychologist William Stern in 1912. To
compute IQ, Stern divided mental age by the actual, chronological age of the person taking
the test and then multiplied by 100 to get rid of the decimal point. For example, if a 6-year-
old girl scored a mental age of 9, she would be assigned an IQ of 150 (9/6 × 100). If a 12-
year-old boy scored a mental age of 6, he would be given an IQ of 50 (6/12 × 100). The IQ
score, as originally computed, expressed a person's mental age relative to his or her
chronological age. Although this formula works adequately for comparing children, it does
not work well for adults because intelligence levels off during adulthood. For example, a 40-
year-old person who scored the same as the average 20-year-old would have an IQ of only
50.

Modern intelligence tests—including the current Stanford-Binet test—no longer compute

scores using the IQ formula. Instead, intelligence tests give a score that reflects how far the
person's performance deviates from the average performance of others who are the same
age. Most modern tests arbitrarily define the average score as 100. By convention, many
people still use the term IQ to refer to a score on an intelligence test.

D. Creation of Group Tests

Print section

During World War I (1914-1918) a group of American psychologists led by Robert M. Yerkes
offered to help the United States Army screen recruits using intelligence tests. Yerkes and
his colleagues developed two intelligence tests: the Army Alpha exam for literate recruits,
and the Army Beta exam for non-English speakers and illiterate recruits. Unlike previous
intelligence tests, which required an examiner to test and interact with each person
individually, the Army Alpha and Beta exams were administered to large groups of recruits
at the same time. The items on the tests consisted of practical, short-answer problems. The
Alpha exam included arithmetic problems, tests of practical judgment, tests of general
knowledge, synonym-antonym comparisons, number series problems, analogies, and other
problems. The Beta exam required recruits to complete mazes, complete pictures with
missing elements, recognize patterns in a series, and solve other puzzles. The army
assigned letter grades of A through D- based on how many problems the recruit answered
correctly. The army considered the highest-scoring recruits as candidates for officer training
and rejected the lowest-scoring recruits from military service. By the end of World War I,
psychologists had given intelligence tests to approximately 1.7 million recruits. Modern
critics have pointed out that the army tests were often improperly administered. For
example, different test administrators used different standards to determine which recruits
were illiterate and should be assigned to take the nonverbal Beta exam. Thus, some recruits
mistakenly assigned to the Alpha exam may have scored poorly because of their limited
English skills, not because of low intelligence.

The use of intelligence tests by the United States military enhanced the credibility and
visibility of group mental tests. Following World War I these tests grew in popularity. Most
were short-answer tests modeled on the army tests or the Stanford-Binet. For example,
Yerkes and Terman developed the National Intelligence Test, a group test for schoolchildren,
around 1920. The Scholastic Aptitude Test, or SAT, was introduced in 1926 as a multiple-
choice exam to aid colleges and universities in their selection of prospective students.

E. Modern Intelligence Tests

Print section

The most widely used modern tests of intelligence are the Stanford-Binet, the Wechsler
Intelligence Scale for Children (WISC), the Wechsler Adult Intelligence Scale (WAIS), and
the Kaufman Assessment Battery for Children (Kaufman-ABC). Each of the tests consists of
a series of 10 or more subtests. Subtests are sections of the main test in which all of the
items are similar. Examples of subtests include vocabulary (“Define happy”), similarities
(“In what way are an apple and pear alike?”), digit span (repeating digit strings of
increasing length from memory), information (“Who was the first president of the United
States?”), object assembly (putting together puzzles), mazes (tracing a path through a
maze), and simple arithmetic problems. Each item has scoring criteria so the examiner can
determine if the answer given is correct.

Items on each subtest are given in order of difficulty until the person being tested misses a
certain number of items. Each subtest provides a score. The subtest scores are then added
together to obtain a total raw score, which is then converted into an IQ score. Some tests,
such as the Wechsler tests, give separate verbal and performance (nonverbal) scores as
well as an overall score.

Other intelligence tests, like the Peabody Picture Vocabulary Test or Raven's Progressive
Matrices, consist of only one item type. In the Peabody Picture Vocabulary Test, the test
taker must define a word by deciding which picture out of four pictures best represents the
meaning of the word said by the examiner. In Raven's Progressive Matrices, a person is
shown a matrix of patterns with one pattern missing. The person must figure out the rules
governing the patterns and then use these rules to pick the item that best fills in the
missing pattern. The Raven's test was designed to minimize the influence of culture by
relying on nonverbal problems that require abstract reasoning and do not require knowledge
of a particular culture.
All of the tests mentioned so far can be individually administered. An examiner tests one
person at a time for a specific amount of time, ranging from 20 to 90 minutes. There are
also group-administered tests. The Army Alpha test described above was one of the earliest
group-administered tests. This test developed into what is now known as the Armed
Services Vocational Achievement Battery (ASVAB), which is used to select and classify
military recruits. Group tests usually are not as reliable as individually administered tests.
They are often shorter and have less variety in item types because of restrictions inherent in
group administration. Furthermore, the administrator of an individual test can more fully
supervise the test taker's performance. For example, the administrator can make sure the
test taker is motivated and provide additional information when necessary. But group tests
are efficient because they can be given to large numbers of people in a short time and at a
relatively low cost.

Achievement tests and aptitude tests are very similar to intelligence tests. An achievement
test is designed to assess what a person has already learned, whereas an aptitude test is
designed to predict future performance or assess potential for learning. Usually the items on
achievement tests and aptitude tests relate to a specific area of knowledge, such as
mathematics or vocabulary. Because intelligence tests frequently include these same areas
of knowledge, many experts believe that it is impossible to distinguish between intelligence
tests, achievement tests, and aptitude tests. Often, test makers call their tests achievement
tests or aptitude tests to avoid the word intelligence, which can be frightening to some test
takers. Examples of achievement and aptitude tests that are widely used include the SAT,
the Graduate Record Exam (GRE), the California Achievement Test, the Law School
Admissions Test (LSAT), and the Medical College Admission Test (MCAT).

F. Standardization, Reliability, and Validity of Tests

Print section

An intelligence test, like any other psychological test, must meet certain criteria in order to
be accepted as scientific and accurate. A test must be standardized, reliable, and valid.

Standardization refers to the process of defining norms of performance to which all test
takers are compared. Before an intelligence test can be used to make meaningful
comparisons, the test makers first give the test to a sample of the population representative
of the individuals for whom the test is designed. This sample of people is called a normative
sample, because it is used to establish norms (standards) of performance on the test.
Normative samples usually consist of thousands of people from all areas of the country and
all strata of society. Test scores of people in the sample are statistically analyzed to compile
the test norms. When the test is made available for general use, these norms are used to
determine a score for each person who takes the test. The IQ score or overall score reflects
how well the person did compared to people of the same age in the normative sample.

Reliability refers to the consistency of test scores. A reliable test yields the same or close to
the same score for a person each time it is administered. In addition, alternate forms of the
test should produce similar results. By these criteria, modern intelligence tests are highly
reliable. In fact, intelligence tests are the most reliable of all psychological tests.

Validity is the extent to which a test predicts what it is designed to predict. Intelligence
tests were designed to predict school achievement, and they do that better than they do
anything else. For example, IQ scores of elementary school students correlate moderately
with their class grades and highly with achievement test scores. IQ tests also predict well
the number of years of education that a person completes. The SAT is somewhat less
predictive of academic performance in college. Educators note that success in school
depends on many other factors besides intelligence, including encouragement from parents
and peers, interest, and motivation.

Intelligence tests also correlate with measures of accomplishment other than academic
success, such as occupational status, income, job performance, and other measures of
vocational success. However, IQ scores do not predict occupational success as well as they
predict academic success. Twenty-five percent or less of the individual differences in
occupational success are due to IQ. Therefore, a substantial portion of the variability in
occupational success—75 percent or more—is due to factors other than intelligence.

Validity also refers to the degree to which a test measures what it is supposed to measure.
A valid intelligence test should measure intelligence and not some other capability.
However, making a valid intelligence test is not a straightforward task because there is little
consensus on a precise definition of intelligence. Lacking such a consensus, test makers
usually evaluate validity by determining whether test performance correlates with
performance on some other measure assumed to require intelligence, such as achievement
in school.

G. Distribution of IQ Scores

Print section

IQ scores, like many other biological and psychological characteristics, are distributed
according to a normal distribution, which forms a normal curve, or bell curve, when plotted
on a graph. In a normal distribution, most values fall near the average, and few values fall
far above or far below the average. Although raw scores are not exactly normally
distributed, test makers derive IQ scores using a formula that forces the scores to conform
to the normal distribution. The normal distribution is defined by its mean (average score)
and its standard deviation (a measure of how scores are dispersed relative to the mean).
Usually the mean of an IQ test is arbitrarily set at 100 with a standard deviation of 15.
Other tests use different values. For example, the SAT originally used a mean of 500 and a
standard deviation of 100, although these are now recomputed annually.

Because IQs are distributed along a normal curve, a fixed percentage of scores fall between
the mean and any standard deviation value. For example, 34 percent of IQ scores fall
between the mean and one standard deviation. For a standard IQ distribution with a mean
of 100 and a standard deviation of 15, 34 percent of the cases would fall between 100 and
115. Since the normal curve is symmetrical about the mean, 34 percent of the scores would
also fall between 85 and 100, which represents one standard deviation below the mean. To
interpret the score of any test, it is important to know the mean and standard deviation of
the test. Along with knowledge of the standard deviation and the normative sample used for
the test, one can then interpret the score in terms of the percentage of the population
scoring higher or lower. If a person obtains a score of 115 on an IQ test, approximately 16
percent of the population will score higher and 84 percent will score lower.

When an IQ test is revised, it is restandardized with a new normative sample. The

distribution of raw scores in the sample population determines the IQ that will be assigned
to the raw scores of others who take the test. By analyzing the performance over the years
of different normative samples on the same tests, researchers have concluded that
performance on intelligence tests has risen significantly over time. This phenomenon,
observed in industrialized countries around the world, is known as the Flynn effect, named
after the researcher who discovered it, New Zealand philosopher James Flynn. Scores on
some tests have increased dramatically. For example, scores on the Raven's Progressive
Matrices, a widely used intelligence test, increased 15 points in 50 years when scored by the
same norms. In other words, a representative sample of the population that took the test in
1992 scored an average of 15 points higher on the test than a representative sample that
took the test in 1942.

It appears that people are getting smarter. However, only some tests show these changes.
Tests of visual-spatial reasoning, like the Raven's test, show the largest changes, while
vocabulary and verbal tests show almost no change. Some psychologists believe that people
are not really getting smarter but are only becoming better test takers. Others believe the
score gains reflect real increases in intelligence and speculate they may be due to improved
nutrition, better schooling, or even the effects of television and video games on visual-
spatial reasoning.

H. Uses of Intelligence Tests

Print section

Intelligence tests and similar tests are widely used in schools, business, government, the
military, and medicine. In many cases, intelligence tests are used to avoid the biases more
arbitrary methods of selection introduce. For example, it was once common for colleges to
admit students whose parents had attended the college or who came from socially
prominent families. By using tests, colleges could select students based on their ability
instead of their social position.

Intelligence tests were originally designed for use in schools. In elementary and secondary
schools, educators use tests to assess how well a student can be expected to perform and
to determine if special educational programs are necessary. Intelligence tests can help to
identify students with mental retardation and to determine an appropriate educational
program for these students (see Education of Students with Mental Retardation).
Intelligence tests may also be required for admission into programs for the gifted or
talented (see Education of Gifted Students). Institutions of higher education use
achievement or aptitude tests, which are very similar to intelligence tests, for the selection
and placement of students.

In business, employers frequently use intelligence and aptitude tests to select job
applicants. Since World War I, the United States military has had one of the most
comprehensive testing programs for selection and job assignment. Anyone entering the
military takes a comprehensive battery of tests, including an intelligence test. For
specialized and highly skilled jobs in the military, such as jet pilot, the testing is even more
rigorous. Intelligence tests are helpful in the selection of individuals for complex jobs
requiring advanced skills. The major reason intelligence tests work in job selection is that
they predict who will learn new information required for the job. To a lesser extent, they
predict who will make “smart” decisions on the job.

In medicine, physicians use intelligence tests to assess the cognitive functioning of patients,
such as those with brain damage or degenerative diseases of the nervous system.
Psychiatrists and psychologists may use intelligence tests to diagnose the mental capacities
of their clients.

I. Criticisms of Intelligence Tests

Print section

Properly used, intelligence tests can provide valuable diagnostic information and insights
about intellectual ability that might otherwise be overlooked or ignored. In many
circumstances, however, intelligence testing has become extremely controversial, largely
because of misunderstandings about how to interpret IQ scores.

1. Validity

Print section

One criticism of intelligence tests is that they do not really measure intelligence but only a
narrow set of mental capabilities. For example, intelligence tests do not measure wisdom,
creativity, common sense, social skills, and practical knowledge—abilities that allow people
to adapt well to their surroundings and solve daily problems. The merit of this criticism
depends on how one defines intelligence. Some theorists consider wisdom, creativity, and
social competence aspects of intelligence, but others do not. Psychologists know little about
how to objectively measure these other abilities. Another criticism of IQ tests is that some
people may not perform well because they become anxious when taking any timed,
standardized test. Their poor performance may reflect their anxiety rather than their true
abilities. However, test anxiety is probably not a major cause of incorrect scores.

2. Misinterpretation and Misuse

Print section

Critics of intelligence testing argue that IQ tests tend to be misinterpreted and misused.
Because IQ tests reduce intelligence to a single number, many people mistakenly regard IQ
as if it were a fixed, real trait such as height or weight, rather than an abstract concept that
was originally designed to predict performance in school. Furthermore, some people view IQ
as a measurement of a person's intrinsic worth or potential, even though many factors other
than those measured by IQ tests contribute to life success.

Critics also note that intelligence testing on a large scale can have dangerous social
consequences when the results are misused. For example, during the 1920s IQ tests were
used to identify “feeble-minded” persons. These persons were then subject to forced
sterilization. In the 1927 case Buck v. Bell, the United States Supreme Court upheld the
right of states to sterilize individuals judged to be feeble-minded.

In judging the uses of intelligence tests, one must compare how decisions would be made
without using the tests. When tests are used to make a decision, there should be evidence
that the decision made using the test is better with the test than without it. For example, if
schools did not use intelligence or aptitude tests to determine which students need remedial
education, teachers would be forced to rely on more subjective and unreliable criteria, such
as their personal opinions.

In some cases, institutions use tests when they do not need to. Some colleges and
universities require students to take admission tests but then admit 80 percent or more of
applicants. Tests are of little use in selection decisions when there is little or no selection.
Another criticism of intelligence tests is that they sometimes lead to inflexible cutoff rules.
In some states, for example, a person with mental retardation must have an IQ of 50 or
below before being allowed to work in a special facility known as a sheltered workshop.
Although intelligence is important in determining performance, it is not the only
determinant. People with an IQ of 50 vary widely in their skills and abilities. Using an
arbitrary cutoff of 50 can make it difficult for people whose IQ is 51 to get essential
services.

3. Bias

Print section

Psychologists have long known that ethnic and racial groups differ in their average scores
on intelligence tests. For example, African Americans as a group consistently average 15
points lower than whites on IQ tests. Such differences between groups have led some
people to believe that intelligence tests are culturally biased. Many kinds of test items
appear to require specialized information that might be more familiar to some groups than
to others. Defenders of IQ tests argue, however, that these same ethnic group differences
appear on test items in which cultural content has been reduced.

The question of bias in tests has led intelligence researchers to define bias very precisely
and find ways of explicitly assessing it. An intelligence test free of bias should predict
academic performance equally well for African Americans, Hispanics, whites, men, women,
and any other subgroups in the population. Based on this definition of bias, experts agree
that intelligence tests in wide use today have little or no bias for any groups that have been
assessed. Many psychologists believe that group differences in performance exist not
because of inherent flaws in the tests, but because the tests merely reflect social and
educational disadvantages experienced by members of certain racial and ethnic groups in
school and other settings. For more information on IQ differences between groups, see the
Racial and Ethnic Differences section of this article.

Because they are used for educational and employment testing, tests have been challenged
in many court cases. In the 1979 case Larry P. v. Wilson Riles, a group of black parents in
California argued that intelligence tests were racially biased. As evidence they cited the fact
that black children were disproportionately represented in special education classes.
Placement in these classes depended in part on the results of IQ tests. A federal judge
hearing the case concluded that the tests were biased and should not be used to place black
children in special education. The judge also ordered the state of California to monitor and
eliminate disproportionate placement of black children in special education classes. In a
1980 case, PASE v. Hanon, brought in Chicago on the same grounds, a federal judge ruled
that the IQ tests being used were not biased (except for a few items). In employment
cases, a number of rulings have specified how tests can be used. For example, it is not legal
to test applicants for an ability that is not required to do the job.
IV. Theories of Intelligence

Print section

Scholars have tried to understand the nature of intelligence for many years, but they still do
not agree on a single theory or definition. Some theorists try to understand intelligence by
analyzing the results of intelligence tests and identifying clusters of abilities. Other theorists
believe that intelligence encompasses many abilities not captured by tests. In recent years,
some psychologists have tried to explain intelligence from a biological standpoint.

A. General Intelligence

Print section

Efforts to explain intelligence began even before Binet and Simon developed the first
intelligence test. In the early 1900s British psychologist Charles Spearman made an
important observation that has influenced many later theories of intelligence: He noted that
all tests of mental ability were positively correlated. Correlation is the degree to which two
variables are associated and vary together (see Psychology: Correlational Studies).
Spearman found that individuals who scored high on any one of the mental tests he gave
tended to score high on all others. Conversely, people who scored low on any one mental
test tended to score low on all others.

Spearman reasoned that if all mental tests were positively correlated, there must be a
common variable or factor producing the positive correlations. In 1904 Spearman published
a major article about intelligence in which he used a statistical method to show that the
positive correlations among mental tests resulted from a common underlying factor. His
method eventually developed into a more sophisticated statistical technique known as factor
analysis. Using factor analysis, it is possible to identify clusters of tests that measure a
common ability.

Based on his factor analysis, Spearman proposed that two factors could account for
individual differences in scores on mental tests. He called the first factor general intelligence
or the general factor, represented as g. According to Spearman, g underlies all intellectual
tasks and mental abilities. The g factor represented what all of the mental tests had in
common. Scores on all of the tests were positively correlated, Spearman believed, because
all of the tests drew on g. The second factor Spearman identified was the specific factor, or
s. The specific factor related to whatever unique abilities a particular test required, so it
differed from test to test. Spearman and his followers placed much more importance on
general intelligence than on the specific factor.

Throughout his life, Spearman argued that g, as he had mathematically defined it using
factor analysis, was really what scientists should mean by intelligence. He was also aware
that his mathematical definition of general intelligence did not explain what produced g. In
the 1920s he suggested that g measured a mental “power” or “energy.” Others who have
continued to investigate g speculate that it may relate to neural efficiency, neural speed, or
some other basic properties of the brain.

B. Primary Mental Abilities

Print section

Much of the research on mental abilities that followed Spearman consisted of challenges to
his basic position. In the early 20th century, a number of psychologists produced
alternatives to Spearman's two-factor theory by using different methods of factor analysis.
These researchers identified group factors, specific abilities thought to underlie particular
groups of test items. For example, results from tests of vocabulary and similarities (“How
are an apple and orange alike?”) tend to correlate with each other but not with tests of
spatial ability. Both the vocabulary and similarities tests contain verbal content, so
psychologists might identify a verbal factor based on the correlation between the tests.
Although most psychologists agreed that specialized abilities or group factors existed, they
debated the number of factors and whether g remained as an overall factor.

In 1938 American psychologist Louis L. Thurstone proposed that intelligence was not one
general factor, but a small set of independent factors of equal importance. He called these
factors primary mental abilities. To identify these abilities, Thurstone and his wife, Thelma,
devised a set of 56 tests. They administered the battery of tests to 240 college students and
analyzed the resulting test scores with new methods of factor analysis that Thurstone had
devised. Thurstone identified seven primary mental abilities: (1) verbal comprehension, the
ability to understand word meanings; (2) verbal fluency, or speed with verbal material, as in
making rhymes; (3) number, or arithmetic, ability; (4) memory, the ability to remember
words, letters, numbers, and images; (5) perceptual speed, the ability to quickly distinguish
visual details and perceive similarities and differences between pictured objects; (6)
inductive reasoning, or deriving general ideas and rules from specific information; and (7)
spatial visualization, the ability to mentally visualize and manipulate objects in three
dimensions.

Others who reanalyzed Thurstone's results found two problems with his conclusions. First,
Thurstone used only college students as subjects in his research. College students perform
better on intelligence tests than do individuals in the general population, so Thurstone's
subjects did not represent the full range of intellectual ability. By restricting the range of
ability in his sample, he drastically reduced the size of the correlations between tests. These
low correlations contributed to his conclusion that no general intelligence factor existed. To
understand why restricting the range of ability reduces the size of correlations, consider an
analogy. Most people would agree that in basketball, height is important in scoring. But in
the National Basketball Association (NBA), the correlation between players' scoring and
heights is zero. The reason is that NBA players are heavily selected for their height and
average 15 cm (6 in) taller than the average height in the general population. When
Thurstone gave his tests to a more representative sample of the population, he found larger
correlations among his tests than he had found using only college students.

A second problem with Thurstone's results was that, even in college students, the tests that
Thurstone used were still correlated. The method of factor analysis that Thurstone had
devised made the correlations harder to identify. When other researchers reanalyzed his
data using other methods of factor analysis, the correlations became apparent. The
researchers concluded that Thurstone's battery of tests identified the same g factor that
Spearman had identified.

C. Fluid Intelligence and Crystallized Intelligence

Print section

In the 1960s American psychologists Raymond Cattell and John Horn applied new methods
of factor analysis and concluded there are two kinds of general intelligence: fluid intelligence
(gf) and crystallized intelligence (gc). Fluid intelligence represents the biological basis of
intelligence. Measures of fluid intelligence, such as speed of reasoning and memory,
increase into adulthood and then decline due to the aging process. Crystallized intelligence,
on the other hand, is the knowledge and skills obtained through learning and experience. As
long as opportunities for learning are available, crystallized intelligence can increase
indefinitely during a person's life. For example, vocabulary knowledge is known to increase
in college professors throughout their life span.

In addition to identifying the two subtypes of general intelligence, Cattell also developed
what he called investment theory. This theory sought to explain how an investment of
biological endowments (fluid intelligence) could contribute to learned skills and knowledge
(crystallized intelligence). As one might expect, it is very difficult to separate the biological
basis of intelligence from what is learned. As Cattell was aware, nearly all mental tests draw
on both crystallized and fluid intelligence. Consequently, crystallized and fluid abilities are
correlated with each other. Some researchers interpret this correlation between the two
factors as evidence of Spearman's factor of general intelligence, g. They see Cattell's theory
as a refinement of Spearman's original theory, not a departure from it.

D. Multiple Intelligences

Print section

In 1983 American psychologist Howard Gardner proposed a theory that sought to broaden
the traditional definition of intelligence. He felt that the concept of intelligence, as it had
been defined by mental tests, did not capture all of the ways humans can excel. Gardner
argued that we do not have one underlying general intelligence, but instead have multiple
intelligences, each part of an independent system in the brain.

In formulating his theory, Gardner placed less emphasis on explaining the results of mental
tests than on accounting for the range of human abilities that exist across cultures. He drew
on diverse sources of evidence to determine the number of intelligences in his theory. For
example, he examined studies of brain-damaged people who had lost one ability, such as
spatial thinking, but retained another, such as language. The fact that two abilities could
operate independently of one another suggested the existence of separate intelligences.
Gardner also proposed that evidence for multiple intelligences came from prodigies and
savants. Prodigies are individuals who show an exceptional talent in a specific area at a
young age, but who are normal in other respects. Savants are people who score low on IQ
tests—and who may have only limited language or social skills—but demonstrate some
remarkable ability, such as extraordinary memory or drawing ability. To Gardner, the
presence of certain high-level abilities in the absence of other abilities also suggested the
existence of multiple intelligences.

Gardner initially identified seven intelligences and proposed a person who exemplified each
one. Linguistic intelligence involves aptitude with speech and language and is exemplified by
poet T. S. Eliot. Logical-mathematical intelligence involves the ability to reason abstractly
and solve mathematical and logical problems. Physicist Albert Einstein is a good example of
this intelligence. Spatial intelligence is used to perceive visual and spatial information and to
conceptualize the world in tasks like navigation and in art. Painter Pablo Picasso represents
a person of high spatial intelligence. Musical intelligence, the ability to perform and
appreciate music, is represented by composer Igor Stravinsky. Bodily-kinesthetic
intelligence is the ability to use one's body or portions of it in various activities, such as
dancing, athletics, acting, surgery, and magic. Martha Graham, the famous dancer and
choreographer, is a good example of bodily-kinesthetic intelligence. Interpersonal
intelligence involves understanding others and acting on that understanding and is
exemplified by psychiatrist Sigmund Freud. Intrapersonal intelligence is the ability to
understand one's self and is typified by the leader Mohandas Gandhi. In the late 1990s
Gardner added an eighth intelligence to his theory: naturalist intelligence, the ability to
recognize and classify plants, animals, and minerals. Naturalist Charles Darwin is an
example of this intelligence. According to Gardner, each person has a unique profile of these
intelligences, with strengths in some areas and weaknesses in others.

Gardner's theory found rapid acceptance among educators because it suggests a wider goal
than traditional education has adopted. The theory implies that traditional school training
may neglect a large portion of human abilities, and that students considered slow by
conventional academic measures might excel in other respects. A number of schools have
formed with curriculums designed to assess and develop students' abilities in all of the
intelligences Gardner identified.

Critics of the multiple intelligences theory have several objections. First, they argue that
Gardner based his ideas more on reasoning and intuition than on empirical studies. They
note that there are no tests available to identify or measure the specific intelligences and
that the theory largely ignores decades of research that show a tendency for different
abilities to correlate—evidence of a general intelligence factor. In addition, critics argue that
some of the intelligences Gardner identified, such as musical intelligence and bodily-
kinesthetic intelligence, should be regarded simply as talents because they are not usually
required to adapt to life demands.

E. Triarchic Theory of Intelligence

Print section

In the 1980s American psychologist Robert Sternberg proposed a theory of intelligence that,
like Gardner's theory of multiple intelligences, attempted to expand the traditional
conception of intelligence. Sternberg noted that mental tests are often imperfect predictors
of real-world performance or success. People who do well on tests sometimes do not do as
well in real-world situations. According to Sternberg's triarchic (three-part) theory of
intelligence, intelligence consists of three main aspects: analytic intelligence, creative
intelligence, and practical intelligence. These are not multiple intelligences as in Gardner's
theory, but interrelated parts of a single system. Thus, many psychologists regard
Sternberg's theory as compatible with theories of general intelligence.

Analytic intelligence is the part of Sternberg's theory that most closely resembles the
traditional conception of general intelligence. Analytic intelligence is skill in reasoning,
processing information, and solving problems. It involves the ability to analyze, evaluate,
judge, and compare. Analytic intelligence draws on basic cognitive processes or
components.

Creative intelligence is skill in using past experiences to achieve insight and deal with new
situations. People high in creative intelligence are good at combining seemingly unrelated
facts to form new ideas. According to Sternberg, traditional intelligence tests do not
measure creative intelligence, because it is possible to score high on an IQ test yet have
trouble dealing with new situations.

Practical intelligence relates to people's ability to adapt to, select, and shape their real-world
environment. It involves skill in everyday living (“street smarts”) and in adapting to life
demands, and reflects a person's ability to succeed in real-world settings. An example given
by Sternberg of practical intelligence is of an employee who loved his job but hated his
boss. An executive recruiter contacted the employee about a possible new job. Instead of
applying for the job, the employee gave the recruiter the name of his boss, who was
subsequently hired away from the company. By getting rid of the boss he hated instead of
leaving the job he loved, the employee showed adaptation to his real-world environment.
People with high practical intelligence may or may not perform well on standard IQ tests.

In Sternberg's view, “successfully intelligent” people are aware of their strengths and
weaknesses in the three areas of intelligence. They figure out how to capitalize on their
strengths, compensate for their weaknesses, and further develop their abilities in order to
achieve success in life.

Sternberg's theory has drawn praise because it attempts to broaden the domain of
intelligence to more exactly correspond to what people frequently think intelligence is. On
the other hand, some critics believe that scientific studies do not support Sternberg's
proposed triarchic division. For example, some propose that practical intelligence is not a
distinct aspect of intelligence, but a set of abilities predicted by general intelligence.

F. Other Approaches

Print section

Many researchers have taken new approaches to understanding intelligence based on

advances in the neurological, behavioral, and cognitive sciences. Some studies have found
that differences in IQ correspond with various neurological measures. For example, adults
with higher IQs tend to show somewhat different patterns of electrical activity in the brain
than do people with lower IQs. In addition, PET (positron emission tomography) scans show
that adults with higher IQs have lower rates of metabolism for cortical glucose as they work
on relatively difficult reasoning problems than people with lower IQs. That is, people with
higher IQs seem to expend less energy in solving difficult problems than those with lower
IQs. Other researchers have sought to understand human intelligence by using the
computer as a metaphor for the mind and studying how artificial intelligence computer
programs relate to human information processing. These new approaches are extremely
promising, but their ultimate value has yet to be determined.

In recent years a number of theorists have proposed the existence of emotional intelligence
that is complementary to the type of intelligence measured by IQ tests. American
psychologists Peter Salovey and John Mayer, who together introduced the concept in 1990,
define emotional intelligence as the ability to perceive, understand, express, and regulate
emotions. Emotionally intelligent people can use their emotions to guide thoughts and
behavior and can accurately read others' emotions. Daniel Goleman, an American author
and journalist, popularized the concept in his book Emotional Intelligence (1995). He
expanded the concept to include general social competence.
An American psychologist, Douglas Detterman, has compared general intelligence to a
complex system, like a university, city, or country. In this view, IQ tests provide a global
rating reflective of the many cognitive processes and learning experiences that compose
intelligence, just as a rating of a university is based on an evaluation of its components,
such as library size, faculty quality, and size of endowment. Mental tests tend to correlate
with each other because they are part of a unified system that works together. The
implication of this theory is that understanding general intelligence will require
understanding how the cognitive processes of the brain actually work.

V. Influence of Heredity and Environment

Print section

Few topics in the social sciences have produced more controversy than the relative
influences of nature and nurture on intelligence. Is intelligence determined primarily by
heredity or by one's environment?

The issue has aroused intense debate because different views on the heritability of
intelligence lead to different social and political implications. The strictest adherents of a
genetic view of intelligence believe that every person is born with a fixed amount of
intelligence. They argue that there is little one can do to improve intelligence, so special
education programs should not be expected to produce increases in IQ. On the other hand,
those who see intelligence as determined mostly by environmental factors see early
intervention programs as critical to compensate for the effects of poverty and other
disadvantages. In their view, these programs help to create equal opportunities for all
people. Perhaps the most controversial issue surrounding intelligence has been the
assertion by some people that genetic factors are responsible not only for differences in IQ
between individuals, but also for differences between groups. In this view, genetic factors
account for the poorer average performance of certain racial and ethnic groups on IQ tests.
Others regard genetic explanations for group differences as scientifically indefensible and
view as racist the implication that some racial groups are innately less intelligent than
others.

Today, almost all scientists agree that intelligence arises from the influence of both genetic
and environmental factors. Careful study is required in order to attribute any influence to
either environment or heredity. For example, one measure commonly used to assess a
child's home environment is the number of books in the home. But having many books in
the home may be related to the parents' IQ, because highly intelligent people tend to read
more. The child's intelligence may be due to the parents' genes or to the number of books
in the home. Further, parents may buy more books in response to their child's genetically
influenced intelligence. Which of these possibilities is correct cannot be determined without
thorough studies of all the factors involved.

A. Genetic Influences

Print section

In behavioral genetics, the heritability of a trait refers to the proportion of the trait's
variation within a population that is attributable to genetics. The heritability of intelligence is
usually defined as the proportion of the variation in IQ scores that is linked to genetic
factors. To estimate the heritability of intelligence, scientists compare the IQs of individuals
who have differing degrees of genetic relationship. Scientists have conducted hundreds of
studies, involving tens of thousands of participants, that have sought to measure the
heritability of intelligence. The generally accepted conclusion from these studies is that
genetic factors account for 40 to 80 percent of the variability in intelligence test scores, with
most experts settling on a figure of approximately 50 percent. But heritability estimates
apply only to populations and not to individuals. Therefore, one can never say what
percentage of a specific individual's intelligence is inherited based on group heritabilities
alone.

Although any degree of genetic relationship can and has been studied, studies of twins are
particularly informative. Identical twins develop from one egg and are genetically identical
to each other. Fraternal twins develop from separate eggs and, like ordinary siblings, have
only about half of their genes in common. Comparisons between identical and fraternal
twins can be very useful in determining heritability. Scientists have found that the IQ scores
of identical twins raised together are remarkably similar to each other, while those of
fraternal twins are less similar to each other. This finding suggests a genetic influence in
intelligence. Interestingly, fraternal twins' IQ scores are more similar to each other than
those of ordinary siblings, a finding that suggests environmental effects. Some researchers
account for the difference by noting that fraternal twins are probably treated more alike
than ordinary siblings because they are the same age.

Some of the strongest evidence for genetic influences in intelligence comes from studies of
identical twins adopted into different homes early in life and thus raised in different
environments. Identical twins are genetically identical, so any differences in their IQ scores
must be due entirely to environmental differences and any similarities must be due to
genetics. Results from these studies indicate that the IQ scores of identical twins raised
apart are highly similar—nearly as similar as those of identical twins raised together. For
adoption studies to be valid, placement of twin pairs must be random. If brighter twin pairs
are selectively placed in the homes of adoptive parents with higher intelligence, it becomes
impossible to separate genetic and environmental influences.

Another way of studying the genetic contribution to intelligence is through adoption studies,
in which researchers compare adopted children to their biological and adoptive families.
Adopted children have no genetic relationship to their adoptive parents or to their adoptive
parents' biological children. Thus, any similarity in IQ between the adopted children and
their adoptive parents or the parents' biological children must be due to the similarity of the
environment they all live in, and not to genetics.

There are two interesting findings from studies of adopted children. First, the IQs of adopted
children have only a small relationship to the IQs of their adoptive parents and the parents'
biological children. Second, after the adopted child leaves home, this small relationship
becomes smaller. In general, the IQs of adopted children are always more similar to their
biological parents' IQs than to their adoptive parents' IQs. Further, once they leave the
influence of their adoptive home, they become even more similar to their biological parents.
Both of these findings suggest the importance of hereditary factors in intelligence.

People sometimes assume that if intelligence is highly heritable, then it cannot be changed
or improved through environmental factors. This assumption is incorrect. For example,
height has very high heritability, yet average heights have increased in the 20th century
among the populations of many industrialized nations, most likely because of improved
nutrition and health care. Similarly, performance on IQ tests has increased with each
generation (see the Distribution of IQ Scores section of this article), yet few scientists
attribute this phenomenon to genetic changes. Thus, many experts believe that improved
environments can, to some degree, increase a person's intelligence.

Some genetic disorders, such as phenylketonuria (PKU) and Down syndrome, may result in
mental retardation and low IQ. But evidence for genetic influences should not be interpreted
as evidence of a direct connection between genes and intelligence. In PKU, for example, a
rare combination of recessive genes sets the stage for a series of biochemical interactions
that ultimately results in low IQ. These interactions only occur, however, in the presence of
the amino acid phenylalanine. If the disorder is detected early and phenylalanine is withheld
from the infant's diet, then large IQ deficits do not develop.

B. Environmental Influences

Print section

If genetic influences account for between 40 and 80 percent of the variation in intelligence,
then environmental influences account for between 20 and 60 percent of the total variation.
Environmental factors comprise all the stimuli a person encounters from conception to
death, including food, cultural information, education, and social experiences. Although it is
known that environmental factors can be potent forces in shaping intelligence, it is not
understood exactly how they contribute to intelligence. In fact, scientists have identified few
specific environmental variables that have direct, unambiguous effects on intelligence. Many
environmental variables have small effects and differ in their effect on each person, making
them difficult to identify.

Schooling is an important factor that affects intelligence. Children who do not attend school
or who attend intermittently score more poorly on IQ tests than those who attend regularly,
and children who move from low-quality schools to high-quality schools tend to show
improvements in IQ. Besides transmitting information to students directly, schools teach
problem solving, abstract thinking, and how to sustain attention—all skills required on IQ
tests.

Many researchers have investigated whether early intervention programs can prevent the
lowered intelligence that may result from poverty or other disadvantaged environments. In
the United States, Head Start is a federally funded preschool program for children from
families whose income is below the poverty level. Head Start and similar programs in other
countries attempt to provide children with activities that might enhance cognitive
development, including reading books, learning the alphabet and the numbers, learning the
names of colors, drawing, and other activities. These programs often have large initial
effects on IQ scores. Children who participate gain as much as 15 IQ points compared to
control groups of similar children not in the program. Unfortunately, these gains seem to
last only as long as the intervention lasts. When children from these programs enter school,
their IQ declines to the level of control groups over a period of several years. This has come
to be known as the “fade-out” effect.

Even though early intervention preschool programs do not seem to produce lasting IQ
gains, some studies suggest they may have other positive long-term effects. For example,
the Consortium for Longitudinal Studies reported that participants are less likely to repeat
grades, less likely be placed in remedial classes, and more likely to finish high school than
comparable nonparticipants—even though both groups show about the same levels of
academic achievement. Preschoolers in early intervention programs may also benefit from
improved health and nutrition, and their mothers may sometimes benefit from additional
education that the programs provide. Because a substantial portion of the variation in
intelligence is due to environmental factors, early intervention programs should be able to
produce significant and lasting IQ gains once the specific environmental variables that
influence IQ have been identified. Researchers continue to search for the interventions that
will increase IQ and, ultimately, academic achievement.

Two environmental variables known to affect intelligence are family size and birth order.
Children from smaller families and children who are earlier-born in their families tend to
have higher intelligence test scores. These effects, however, are very small and amount to
only a few IQ points. They are detectable only when researchers study very large numbers
of families.

Although there has been substantial debate about the effects of other environmental
variables, certain substances in the prenatal environment may influence later intelligence.
For example, some pregnant women who consume large amounts of alcohol give birth to
children with fetal alcohol syndrome, a condition marked by physical abnormalities, mental
retardation, and behavioral problems. Even exposure to moderate amounts of alcohol may
have some negative influence on the development of intelligence, and to date no safe
amount of alcohol has been established for pregnant women. Scientists have also
discovered that certain substances encountered during infancy or childhood may have
negative affects on intelligence. For example, children with high blood levels of lead, as a
result of breathing lead-contaminated air or eating scraps of lead-based paint, tend to have
lower IQ scores. Prolonged malnutrition during childhood also seems to influence IQ
negatively. In each of these cases, a correlation exists between environmental factors and
measured intelligence, but one cannot conclude that these factors directly influence
intelligence. Other environmental variables in this category include parenting styles and the
physical environment of the home.

Although the nature-nurture debate has raged for some time, research points to a
conclusion that appeals to common sense: Intelligence is about half due to nature
(heredity) and about half due to nurture (environment). The exact mechanisms by which
genetic and environmental factors operate remain unknown. Identifying the specific
biological and environmental variables that affect intelligence is one of the most important
challenges facing researchers in this field.

C. Sex Differences

Print section

Are women smarter or are men smarter? Psychologists have studied sex differences in
intelligence since the beginning of intelligence testing. The question is a very complicated
one, though. One problem is that test makers sometimes eliminate questions that show
differences between males and females to eliminate bias from the test. Intelligence tests,
therefore, may not show gender differences even if they exist. Even when gender
differences have been explicitly studied, they are hard to detect because they tend to be
small.
There appear to be no substantial differences between men and women in average IQ. But
the distribution of IQ scores is slightly different for men than for women. Men tend to be
more heavily represented at the extremes of the IQ distribution. Men are affected by mental
retardation more frequently than are women, and they also outnumber women at very high
levels of measured intelligence. Women's scores are more closely clustered around the
mean.

Although there are no differences in overall IQ test performance between men and women,
there do seem to be differences in some more specialized abilities. Men, on average,
perform better on tests of spatial ability than do women. Spatial ability is the ability to
visualize spatial relationships and to mentally manipulate objects. The reason for this
difference is unknown. Some psychologists speculate that spatial ability evolved more in
men because men were historically hunters and required spatial ability to track prey and
find their way back from hunting forays. Others believe that the differences result from
parents' different expectations of boys' and girls' abilities.

Many studies have examined whether gender differences exist in mathematical ability, but
the results have been inconsistent. In 1990 American researchers statistically combined the
results of more than 100 studies on gender differences in mathematics using a technique
known as meta-analysis. They found no significant differences in the average scores of
males and females on math tests. Research also indicates that the average girl's grades in
mathematics courses equal or exceed those of the average boy. Other studies have found
that boys and girls perform equally well on math achievement tests during elementary
school, but that girls begin to fall behind boys in later years. For example, male high school
seniors average about 45 points higher on the math portion of the SAT than do females.

A 1995 study examined the performance of more than 100,000 American adolescents on
various mental tests. The study found that on average, females performed slightly better
than males on tests of reading comprehension, writing, perceptual speed, and certain
memory tasks. Males tended to perform slightly better than girls on tests of mathematics,
science, and social studies. In almost all cases, the average sex differences were small.

Are differences in abilities between men and women biologically based or are they due to
cultural influences? There is some evidence on both sides. On the biological side,
researchers have studied androgenized females, individuals who are genetically female but
were exposed to high levels of testosterone, a male hormone, during their gestation. As
these individuals grow up, they are culturally identified as female, but they tend to play with
“boys' toys,” like blocks and trucks, and have higher levels of spatial ability than females
who were not exposed to high levels of testosterone. Further evidence for a biological basis
for spatial gender differences comes from comparisons of the brains of men and women.
Even when corrected for body size, males tend to have slightly larger brains than females.
Some scientists speculate that this extra brain volume in males may be devoted to spatial
ability.

On the cultural side, many social scientists argue that differences in abilities between men
and woman arise from society's different expectations of them and from their different
experiences. Girls do not participate as extensively as boys do in cultural activities thought
to increase spatial and mathematical ability. As children, girls are expected to play with
dolls and other toys that develop verbal and social skills while boys play with blocks, video
games, and other toys that encourage spatial visualization. Later, during adolescence, girls
take fewer math and science courses than boys, perhaps because of stereotypes of math
and science as masculine subjects and because of less encouragement from teachers,
peers, and parents. Many social scientists believe cultural influences account for the
relatively low representation of women in the fields of mathematics, engineering, and the
physical sciences.

It is important to remember that sex differences, where they exist, represent average
differences between men and women as groups, not individuals. Knowing whether an
individual is female or male reveals little about that person's intellectual abilities.

D. Racial and Ethnic Differences

Print section

Numerous studies have found differences in measured IQ between different self-identified

racial and ethnic groups. For example, many studies have shown that there is about a 15-
point IQ difference between African Americans and whites, in favor of whites. The mean
scores of IQ scores of the various Hispanic American subgroups fall roughly midway
between those for blacks and whites. Although these differences are substantial, there are
much larger differences between people within each group than between the means of the
groups. This large variability within groups means that a person's racial or ethnic
identification cannot be used to infer his or her intelligence.

The debate about racial and ethnic differences in IQ scores is not about if the differences
exist but what causes them. In 1969 Arthur Jensen, a psychology professor at the
University of California at Berkeley, ignited the modern debate over racial differences.
Jensen published a controversial article in which he argued that black-white differences in
IQ scores might be due to genetic factors. Further, he argued that if IQ had a substantial
genetic component, remedial education programs to improve IQ should not be expected to
raise IQ as they were currently being applied. In 1994 American psychologist Richard
Herrnstein and American social analyst Charles Murray renewed the debate with the
publication of The Bell Curve (1994). Although only a small portion of the book was devoted
to race differences, that portion of the book received the most attention in the popular
press. Among other arguments, Herrnstein and Murray suggested it was possible that at
least some of the racial differences in average IQ were due to genetic factors. Their
arguments provoked heated debates in academic communities and among the general
public.

As discussed earlier, research supports the idea that differences in measured intelligence
between individuals are partly due to genetic factors. However, psychologists agree that this
conclusion does not imply that genetic factors contribute to differences between groups. No
one knows exactly what causes racial and ethnic differences in IQ scores. Some scientists
maintain that these differences are in part genetically based. Supporters of this view believe
that racial and ethnic groups score differently on intelligence tests partly because of genetic
differences between the groups. Others think the cause is entirely environmental. In this
view, certain racial and ethnic groups do poorer on IQ tests because of cultural and social
factors that put them at a disadvantage, such as poverty, less access to good education,
and prejudicial attitudes that interfere with learning. Representing another perspective,
many anthropologists reject the concept of biological race, arguing that races are socially
constructed categories with little scientific basis (see Race). Because of disagreements
about the origins of group differences in average IQ, conclusions about these differences
must be evaluated cautiously.
Some research indicates that the black-white differential in IQ scores might be narrowing.
Several studies have found that the difference in average IQ scores between African
Americans and whites has shrunk to 10 points or less, although research has not
established this trend clearly. The National Assessment of Educational Progress, a national
longitudinal study of academic achievement, also shows that the performance of African
Americans on math and science achievement tests improved between 1970 and 1996 when
compared to whites.

Educators and researchers have focused much attention on explaining why some ethnic
groups perform more poorly than others on measures of intelligence and academic
achievement. Another topic of research is why some ethnic groups, particularly Asian
Americans, perform so well academically. Compared to other groups, Asian American
students get better grades, score higher on math achievement and aptitude tests, and are
more likely to graduate from high school and college. The exact reasons for their high
academic performance are unknown. One explanation points to Asian cultural values and
family practices that place central importance on academic achievement and link success in
school with later occupational success. Critics counter that this explanation does not explain
why Asian Americans excel in specific kinds of abilities.

The academic and occupational successes of Asian Americans have caused many people to
presume Asian Americans have higher-than-average IQs. However, most studies show no
difference between the average IQ of Asian Americans and that of the general population.
Some studies of Asians in Asia have found a 3 to 7 point IQ difference between Asians and
whites, in favor of Asians, but other studies have found no significant differences.