Intelligence Testing

July 30, 1993


Abstract
By Sarah Glaser

Do traditional IQ tests overlook some bright students?

Ingrained in the nation's educational system, intelligence tests are as American as apple pie. They play a crucial role in everything from college admissions to military job assignments. But IQ and aptitude tests have come under increasing criticism in recent years as measures of real-world intelligence. Although educators still generally consider them good predictors of school performance, a faction of the psychological community is challenging the use of traditional standardized tests. Led by Harvard's Howard Gardner, the upstarts embrace the theory of multiple intelligences, which holds that there are many varieties of intelligence encompassing talents often ignored by traditional IQ tests. As these new theorists exert a growing influence on educators' understanding of intelligence, the nation's classrooms could witness a quiet revolution in teaching and testing.

go to top


Overview

The Jacksonville, Fla., fifth-graders were working intently ontheir projects -- designing cities of the future. Bursting with creative excitement, one boy ran over to a group of educators who were observing the class and described his creation, which featured an elaborate transportation system for the elderly. One of the visitors was so impressed with the boy's detailed explanation that he wondered why he wasn't in the school's gifted and talented (GT) program. But when he checked the boy's test records, he discovered that his IQ score didn't meet the program's minimum.

In Hollywood's version of the story, the boy would have been admitted to the program, and like the Karate Kid, he would have overcome all obstacles and excelled. But this is real life. Although the boy seemed a promising GT candidate, his test scores said otherwise. End of story.

IQ tests, ingrained in the nation's education system, are as American as apple pie. But their ability to distinguish intellectual talent has been questioned since their creation at the turn of the century. Now, iconoclastic researchers are again challenging standardized intelligence tests, once more raising emotional questions about how American society balances merit, talent and equal opportunity in an ethnically and socially diverse nation.

The current wave of concern about intelligence tests -- which include IQ, aptitude and ability tests -- goes back to 1969, when psychologist Arthur R. Jensen set off a firestorm of controversy by suggesting that IQ differences between blacks and whites were due to genetic factors. Jensen's troubling assertion spawned press reports that tended to portray IQ testing as a distasteful and inaccurate measure of individual ability. [1]

But even as the public's faith in IQ tests was diminishing, many experts were embracing the idea that IQ tests accurately measure intelligence. According to one survey, most experts “seem to believe that intelligence tests are doing a good job measuring intelligence, as they would define it.” [2] In the survey of 661 social scientists, 80 percent of the respondents said IQ tests effectively measure at least one important element of intelligence -- abstract reasoning. The survey also showed, however, that many experts feel the tests don't fully measure other key aspects of intelligence, such as creativity.

Jensen's view that IQ is largely inherited is finding wider acceptance. Jensen's position has been reinforced by studies of twins raised apart. [3] (See story, p. 662.) While the findings are uncertain, at best, scientific journals nonetheless are crowded with new studies finding correlations between biological activity in the brain and IQ score.

Today, in most U.S. school districts, students take at least two multiple-choice IQ tests during their early school years. While the tests carry less weight than they did a decade ago, educators still find them useful. High scores can still be the ticket to America's own brand of privileged education -- public school classes for the gifted or expensive private schools geared to high achievers. At the other end of the spectrum, individual IQ tests administered by school psychologists are frequently used to diagnose learning disabilities and mental retardation.

During the Great Society era of the 1960s, the founders of programs like Head Start hoped to raise the IQs of poor children, many of them black, with early education and extra doses of parenting and nutrition. [4] But Arthur Jensen suggested that the gap between the races couldn't be closed by such programs because variability in IQ stemmed mainly from genetic factors.

Today, Jensen, a professor of educational psychology at the University of California-Berkeley, is even more convinced that biology causes the 15-point gap in the average scores of blacks and whites. “I think it's probably largely a genetic difference between the racial groups,” he says.

But even those who credit biology's role in intelligence say it doesn't fully explain the difference between the IQ scores of blacks and whites. If biology explains even half the story, they say, environmental conditions -- inferior schooling, limited exposure to middle-class culture and the low status of blacks in American society -- may well explain the other half.

With so many varieties of intelligence, Gardner says that new ways of testing will help educators find bright children even among those who would have scored just average on traditional intelligence tests. “Where individuals differ,” he argues, “is in the strength of these intelligences.” [5]

Stephen J. Ceci, a professor of developmental psychology at Cornell University, is among the critics of traditional testing. “I believe IQ is an achievement test,” he says, “and like all other achievement tests it's directly sensitive to things kids are learning in school. If you take college-prep math, you're going to do better on college math achievement tests than if you don't. And there's nothing mysterious or magical about that. But once you call something an IQ test, it connotes that it's a measure of some inherent or native ability that's impervious to schooling.”

In a review of 200 studies, Ceci found that IQ scores rise with the amount of time children spend in school, regardless of the quality of schooling. The finding runs counter to the traditional view of IQ as a stable, unchanging measure of intelligence throughout a child's school career. [6] Similarly, suggests Ceci, certain kinds of logical thinking measured by IQ tests may be much more familiar to white children than to black children because of their different cultural backgrounds.

Dissidents like Ceci and Gardner are up against 70 years of data supporting the validity of IQ tests as predictors of academic and job success. IQ scores have forecast students' grades in school, postal workers' accuracy in sorting mail and military recruits' ability to pilot a tank. [7]

Some IQ advocates suspect that Gardner's broader social agenda -- finding disadvantaged students with special talent and reforming schools -- is driving his scientific arguments. “Are we just giving the label ‘intelligence' to aptitudes in order to democratize the notion of intelligence?” asks Linda S. Gottfredson, a professor in the Department of Educational Studies at the University of Delaware. “It's not egalitarian to say some people are smarter than others. That's why I think multiple intelligence is so popular. It allows some people to be smarter in some ways and some people to be smarter in other ways.”

Educators who are testing alternative theories of intelligence in the schools admit they are stumbling through uncharted territory. Indeed, they have yet to produce the hard data to challenge the IQ test's supremacy. All they have are some heartwarming anecdotes about bright children from poor backgrounds who were overlooked by teachers trained in the old definition of intelligence. Nonetheless, they raise intriguing questions about the true nature of intelligence and our ability to discern it in individuals.

As educators, psychologists and social scientists work to make educational tests more accurate tools for predicting success among the nation's increasingly diverse student populace, these are some of the key questions they are asking:

Is IQ testing a valid measure of intelligence?

The current debate over IQ testing is but the latest -- and most sophisticated -- installment of the so-called “nature-nurture” argument. Back in the 1960s, many psychologists believed that environmental changes, such as improved education, nutrition and parenting, could enhance children's intelligence. But the environmentalists lost steam when “nurturing” programs like Head Start failed to raise IQs. Meanwhile, studies of twins, adopted children and the brain added to the evidence that biological and genetic factors -- in other words, nature -- played key roles in determining intelligence.

The mounting biological evidence confronted the environmentalists with a difficult empirical task, says Case Western Reserve University psychologist Douglas K. Detterman, who edits the journal Intelligence. Proving the influence of a person's surroundings is extremely difficult, he notes, because “the environment is so chaotic and diverse.” On the other hand, the intelligence tests marketed to schools still deliver the goods. “The [IQ] tests still predict who will do well in school and who won't,” he says. “The tests still empirically work.”

The rift between psychologists over the value of IQ tests stems from a stark difference in the way they define intelligence: One camp sees a central, dominant intelligence, the other envisions multiple types of intelligence.

The first camp believes that a predominant form of general intelligence determines, in essence, whether a person is “smart.” These IQ traditionalists, who offer evidence that neurological and genetic factors underlie intelligence, belong to what Howard Gardner calls the “tough-minded” wing of intelligence studies. Gardner's own “tender-minded” faction, favoring cultural and social explanations of intelligence, sees intelligence as a multi-faceted phenomenon that cannot be captured by a single IQ snapshot. [8]

The primary evidence put forward for one overarching kind of intelligence is that an individual's scores on various kinds of tests usually correlate positively with one another, no matter what the subject matter: A person who scores above average (or below average) on the verbal part of an IQ test will generally score above average (or below average) on the math part.

When British psychologist Charles Edward Spearman noticed this correlation near the turn of the century, he hypothesized that a general intelligence, measured in common by all the tests, drove these correlations. He dubbed this defining characteristic of intelligence as, for general intelligence.

“It's really a property of the brain,” maintains Arthur Jensen, who has correlated the component of IQ scores with electrical activity in the brain, reaction times in push-button exercises and the rate at which the brain uses glucose during an IQ test.

Similarly, artists and other creative types don't necessarily excel on IQ tests, Sternberg maintains, because they “test a narrow construct. And the results are taken as if it tells you everything you need to know about intelligence.”

Weighing against Sternberg's theory are stacks of military data, some in volumes the size of an urban telephone book, showing positive correlations between the aptitude scores of new recruits and their performance in training and on the job. General intelligence scores -- as opposed to scores of mechanical or spatial aptitude, for example -- have proven the best predictor of job performance in studies of more than 7,000 pilot and navigator trainees. [9]

In short, says Linda Gottfredson, who has analyzed the occupational data collected by the Army and other employers, the more mentally complex the job, the more important IQ scores become in determining job performance. Gottfredson says her studies support the view that practical skills, as defined by Sternberg, do not stand apart from the general intelligence measured by IQ tests. Rather, she argues, general intelligence is the “enabling condition” that gives an individual the judgment necessary to become a successful entrepreneur or business manager. [10]

The predictive power of such tests, while considered strong by some social scientists, is nowhere near perfect. “There are too many people with sky-high IQs who just totally fail in their lives,” Sternberg contends, “and too many people with low test scores that succeed.”

Whether IQ tests can ferret out creative talent is among the hottest issues. Studies from the 1950s found that creative individuals tended to score high in IQ, but Sternberg isn't impressed. He scoffs, “The kinds of measures [of creativity] that were used in the 1950s were, ‘Here's a paper clip. Think of unusual uses for the paper clip.' That's going to correlate highly with IQ tests because that's the same [kind of question] as the IQ test.”

Sternberg has developed his own test. It asks youngsters to demonstrate creative talents with offbeat tasks like writing a TV ad for Brussels sprouts or devising a plan to detect extraterrestrial aliens. Sternberg says he has identified youngsters who score high on creativity but low on the analytical skills tapped by traditional IQ tests. In a pilot study of 60 high school students who took college- level courses at Yale last summer, Sternberg found that youngsters who were identified in tests as either creative or practical performed better when teaching was targeted to their thinking style.

Countering Jensen's biological studies of IQ, Gardner marshals his own neurological evidence to show that intelligence is multifaceted. He points to victims of stroke and brain damage who continue to perform normally in some intellectual areas even after losing the functions usually associated with the damaged part of the brain. In the words of Harvard psychologist Sheldon H. White, current scientific evidence suggests the brain is “not one computer but a set of computers which crosstalk with one another.”

School psychologists who administer IQ tests often downplay their importance. “They're not really tests of intelligence,” says Kevin P. Dwyer, a school psychologist in Montgomery County, Md. “They're tests of a small sample of some of the behaviors people need to learn academic material.”

While IQ tests can be useful in diagnosing learning problems, Dwyer believes that observing a child in the classroom over time may provide an equally accurate profile. Schools rely heavily on IQ tests to determine eligibility for gifted and special-education programs, Dwyer says, because they save time and money over long-term observation. “The painful thing for me is how often people will take the test and all they want is the IQ number, and they don't care how the kid got there. That drives me berserk. These are human beings, and they're basing their future on an IQ number.”

Harvard's White, who is writing a history of developmental psychology, considers IQ tests “archaic” measures, based on the embryonic psychology of the early 1900s. Moreover, he says, the tests have never been modernized to incorporate current scientific understanding of thinking processes. “This monolithic IQ is a relic of a eugenic fantasy that was very strong at the turn of the century, and it has done over the course of this century more and more damage,” he says.

Do IQ and aptitude tests close or open doors to opportunity?

The staunchest supporters of IQ and aptitude tests argue that they reflect merit, not class. In the words of Fortune magazine columnist Daniel Seligman, author of a new book on intelligence testing, “The connection between IQ and achievement has one positive implication: It tells us that people at the top in American life are probably there because they are more intelligent than others -- which is doubtless the way most of us think it should be.” [11]

That kind of statement is exactly what makes IQ testing so potentially divisive, says Sheldon White. “What you're dealing with is a decaying remnant of establishmentarian thinking at the turn of the century,” he retorts.

In fact, the history of intelligence testing has two interwoven and conflicting strands: On the one hand, a school-reform movement early in the century supported tests that would categorize children from varying social and immigrant backgrounds in order to place them at educational levels appropriate to their abilities. However, some of the tests' originators were outspoken eugenicists interested in weeding out what they saw as feeble-minded individuals of inferior racial stock.

“This transformation of the American educational system into something driven by these individual differences is an extraordinary accomplishment. In my judgment, it's comparable to the fall of monarchies and the rise of republics,” Herrnstein observes.

Yet today, critics charge, such standardized tests are having the opposite effect, labeling disadvantaged children as failures prematurely and creating a self-fulfilling prophecy.

In Boston, the process starts when second-graders take a standardized test that entitles qualified students to enroll in “advanced work” classes. Those classes prepare students to compete for admission to Boston's elite “exam schools,” where about 80 percent of students go on to college, compared with 20 percent for other Boston public schools.

“The test becomes a very strong gatekeeper based on race and class,” says Monty Neill, associate director of FairTest, a Cambridge, Mass., advocacy organization critical of standardized testing. “It's quite clear that to a great extent these tests measure family background and previous experience.” IQ tests play the same function in deciding who gets into gifted and talented programs across the country, he adds.

The issue of whether the tests help or hurt children from poor backgrounds was placed in sharp relief by a recent California case involving black students. Ironically, their parents had sued the state because the children had been unable to take an IQ test. California had banned IQ testing of black students after a 1979 court ruling declared that the tests were racially biased and resulted in disproportionately high numbers of blacks being placed in classes for the educable mentally retarded (see p. 660).

Mary Amaya, a Hispanic parent, filed the suit after she received a letter from the school district recommending a special-education assessment for one of her sons. In a postscript, the district added that the boy, whose father is black, could get an IQ test only if Amaya registered him as Hispanic.

Amaya had found the IQ test helpful in diagnosing learning problems for her older son before the ban took place. She considered the new policy discriminatory, said her attorney, Mark Bredemeier, of the Landmark Legal Foundation, a conservative group in Kansas City, Mo. “Minority families are saying, ‘We appreciate your wanting to protect us from ourselves,'” said Bredemeier. “‘All we ask for is the right to make our own decisions on these tests.'”

Last August a U.S. district court ruled that black parents should be permitted to get their children tested. The decision is being appealed by the plaintiffs in the original 1979 case.

Although Bredemeier argues that students needing special education are no longer placed in racially isolated classes, Neill insists that special classes still constitute a dead end. “All the large-scale studies have shown most placements in special education guarantee you'll never get a postsecondary education,” he says. And blacks, according to FairTest, are three times as likely to be placed in special-education classes as whites.

Are IQ and aptitude tests culturally biased?

An African chieftain in exotic regalia chastises a visiting scholar: “You can't build a hut,” goes the cartoon's caption, “you don't know how to find edible roots, you know nothing about predicting the weather -- you do terribly on our IQ test.” [13]

Much like the visiting scholar in the cartoon, minority children taking intelligence tests often find themselves transported to an exotic and incomprehensible world of middle-class symbols and words, critics of such tests believe. (See box, p. 657.)

“Children from different cultural and ethnic groups are not given fair measure by the tests,” says Patricia O'Connell Ross, director of the U.S. Department of Education's Jacob Javits Gifted and Talented Education Program. The Washington-based program has a congressional mandate to broaden the diversity of American children in gifted programs. “We're certainly opposed to the IQ test as a singular means by which you identify kids,” she says.

Because blacks consistently score an average of 15-18 points lower than whites, the question of cultural bias won't go away. Although whites are generally more affluent and better educated than blacks, family income apparently does not account for the difference. When blacks and whites of similar economic backgrounds are compared, the gap is only reduced slightly -- from 15 points to 10 or 12 points. And tests of black and white children from different socioeconomic status (SES) repeatedly find that low-SES white children score as high as high-SES black children. [14]

Though many blacks will score at the top of the IQ spectrum and many whites score at the bottom, the practical significance of the difference lies in how many blacks and whites fall at extreme ends of the spectrum. IQ cutoffs can decide whether one is admitted into the Army for example. The armed forces are required to screen out the lowest 10 percent of the IQ distribution -- those who score below 75. Only one in 20 white Americans scores below the cutoff, compared with one in five black Americans. [15]

In the scientific community, the old debate over IQ-test bias -- Are individual questions racially and ethnically slanted? -- has shifted, notes Michael Feuer, staff director of the National Research Council's new Board on Testing and Assessment (see p. 666)* Psychologists are beginning to ask whether tests are biased toward the analytical thinking characteristic of mainstream test designers. *

For example, an Educational Testing Service (ETS) study found that males score an average of 35 points higher on the math part of the Scholastic Assessment Test (SAT) than females, even when the two sexes get the same grades in their college math courses. [16] Some observers theorize that males solve problems with analytical, deductive kinds of reasoning while females tend to approach them with a more innovative, open-minded flair that multiple-choice tests can't capture.

Since the 1970s, testing companies have routinely screened questions for bias with culturally diverse groups of experts and lay people, according to Barbara Plake, director of the Buros Institute of Mental Measurement at the University of Nebraska. “You tend to see far fewer overtly biased items,” concedes Neill of FairTest. “But to say you've removed the overtly biased items does not mean you've removed the cultural bias on the tests, because you still have questions that assume the norms of being are white and middle-class.”

As for why middle-class black children still score lower than whites of similar background, Neill says, “Our argument is that there's a combined effect of both class and race. Even higher-income blacks are still segregated and still come from cultural-subgroup experiences.”

Further support for this explanation comes from anthropologist John U. Ogbu of the University of California-Berkeley, who has studied IQ scores and school performance of low-status groups around the world. Ogbu points to the Burakamin, the contemporary descendants of Japanese untouchables, who score lower in IQ and perform more poorly in school than mainstream Japanese, even though they are racially identical.

However, when Burakamin families immigrate to the United States, where their historic caste status is unknown, they perform as well as other Asians in school. Similarly, argues Ogbu, the inferior school and IQ performance of blacks can be explained by the limited economic opportunities they perceive for themselves in a society that still relegates them to a low-caste status. [17]

Diehard IQ adherents dismiss such explanations of the black-white gap. “The cultural-bias explanation will not fly,” writes Seligman, whose book concludes that recognizing the power of IQ to predict job performance could boost the nation's gross national product (GNP). “[I]t collides with the fact that blacks do relatively well on the culturally loaded verbal sections of IQ tests; it is the subtests that are virtually devoid of cultural content -- those emphasizing abstract-reasoning ability -- on which blacks do worst.” [18]

Yet the concept of abstractness is “exceedingly ill-wrought,” retorts Ceci of Cornell. He argues that many of the so-called “abstract” questions on IQ tests are in fact skills taught in school.

In addition, Ceci disputes the oft-repeated claim of IQ adherents that a student's IQ, not family background, is the best predictor of future income or job success. In a recent study, Ceci and a colleague analyzed data from Project Talent, a national study by the American Institutes for Research in Palo Alto, Calif., begun in 1959 that followed the progress of 400,000 high school students into their early 30s. They concluded that family social origins and education were more influential in predicting future earnings than IQ. Given a choice between “being born smart or being born rich, one ought to opt for the latter,” the authors write. [19]

go to top

** The National Research Council is the administrative andanalytical branch of the National Academy of Sciences andthe National Academy of Engineering.


Background

The roots of intelligence testing lie in the 19th century, when Charles Darwin's theory of evolution generated fascination with genetics and natural selection. Although Darwin never applied his theories to psychology, his cousin, the English scientist Sir Francis Galton, did. Galton is considered the father of eugenics, the movement devoted to improving humankind through selective breeding.

In Hereditary Genius, his 1869 study of 977 eminent men, Galton argued that the status of great men is due to their natural gifts, which can be traced back through their families. Great human ability, he wrote, “breeds true.” [20] At his laboratory, Galton tested visitors' reaction times, color perception and steadiness of hand in the belief that more-intellectually refined individuals would have keener sensory capacities.

A competing school of thought about the nature of intelligence emerged from the work of French psychologist Albert Binet in the 1890s. Binet argued that mental functioning could only be assessed by looking at more complex abilities, such as judgment, memory and language.

In 1904, France's minister of public instruction asked Binet and his student Theodore Simon to develop a test to identify retarded schoolchildren. The so-called Binet-Simon scale that was developed in 1905 is considered the first reliable intelligence test. *

Binet and Simon revised the test in 1908 to identify children who would profit from further instruction. Although the test was a great success -- and served as the foundation for American IQ testing -- Binet kept working to refine the definition of intelligence throughout the rest of his life. Much like today's multiple-intelligence psychologists, he sought a more sophisticated understanding of mental capacity. [21]

In 1916, Stanford psychologist Lewis Terman published his own revision of the Binet-Simon scale.* Terman's test, known as the Stanford-Binet, is the standard by which all later intelligence tests have been judged. In the opening chapter of his manual for test users, Terman voiced the conflicting impulses -- those of both eugenicist and egalitarian -- that have characterized the history of the intelligence-testing movement. *

Speaking as a eugenicist, he asserted that identifying retarded students would “ultimately result in curtailing the reproduction of feeble-mindedness and in the elimination of an enormous amount of crime, pauperism and industrial inefficiency.” At the same time, Terman the egalitarian said the test would help schools respond appropriately to bright children and to assign all children to the appropriate school grades. [22]

Expanding Use of Intelligence Testing in the U.S.

Terman's revised test marked the beginning of large-scale testing in the United States. It coincided with the explosive growth of public school populations between 1890 and 1915, an influx of immigrant children and the demise of the ungraded one-room schoolhouse. Terman's test appeared to answer the schools' new need to classify children according to their skill levels.

The rapid expansion of intelligence testing was spurred by the Army's interest during World War I. Terman and a group of colleagues were asked to help the Army develop group intelligence tests to determine eligibility and to assign new recruits to jobs. These tests, known as the Alpha and Beta scales, were administered to more than 1 million recruits during World War I. Terman's original test had been administered individually, precluding its use on a massive scale. Arthur Otis, a student of Terman's, developed a paper-and-pencil version, permitting it to be taken by large groups and graded objectively.

After the war, schools and industries administered the new group tests in growing numbers. School reformers supported them as the rational application of science to educating children of differing abilities. By 1921, 2 million American schoolchildren were being tested annually, primarily to determine academic tracking. [23]

Controversies Erupt

But the Army tests also set off the first public controversy about IQ testing. The Army data reported that members of immigrant groups scored lower than native-born Americans and that the most recently arrived immigrants from Southern and Eastern Europe scored lower than those from Northern and Western Europe. Black recruits scored lowest of all. These data were seized upon by a small but vocal group of eugenicists led by Princeton University psychologist Carl Brigham, who argued that the nation needed to improve its breeding stock. Similar arguments figured in congressional debate over legislation to restrict immigration. [24]

In 1922, columnist Walter Lippman launched the first popular attack on IQ testing. In several articles in The New Republic, Lippman argued the tests did not measure innate intelligence or heredity. He warned that the tests could “lead to an intellectual caste system in which the task of education had given way to the doctrine of predestination and infant damnation.” [25] But by the mid-1920s, the controversy had died down, and intelligence testing continued to expand.

During the 1920s, psychologists refined the tests, becoming increasingly sophisticated in their statistical techniques for measuring their reliability and introducing machine-scoring techniques. In 1926, the College Entrance Examination Board developed the Scholastic Aptitude Test, recently renamed the Scholastic Assessment Test, as an “IQ test for college-aged subjects” in response to requests from colleges. The purpose was to identify the “diamond in the rough,” someone who could overcome poor preparation and learn a great deal if admitted. [26]

By 1930, Terman's intelligence and achievement tests had sales of an estimated 2 million copies a year. By World War II, intelligence testing had become firmly ensconced in American society, and some 10 million Army recruits had been classified into jobs using an ability test. Testing surged again in the 1950s, after the Soviet satellite Sputnik raised concerns about America's ability to find and train the rocket scientists of the future.

While the tests rewarded those who were poor but acculturated, such as the first-generation sons of European Jewish immigrants, they continued to relegate those outside mainstream society -- primarily blacks and new immigrants -- to the bottom of the economic ladder, critics of the tests now say. The birth of the civil rights movement in the 1960s raised new questions about whether the tests exacerbated racial inequality.

Court Challenges

In a landmark 1967 case, Hobson v. Hansen, black parents in Washington, D.C., challenged the school system's use of intelligence tests to assign students to academic tracks, charging that the policy was racially biased. The D.C. Circuit Court found that because the test had been calibrated for a white, middle-class group, it was inappropriate to use, and it struck down the city's tracking system. That case marked a new reluctance by public schools to use intelligence tests to make routine educational decisions affecting most children.

The issue arose again in 1971, when the parents of seven black children sued the state of California, claiming that their children had been wrongly assigned to classes for the mentally retarded based on culturally biased IQ tests. In 1979, 9th U.S. District Judge Robert F. Peckham ruled in the Larry P. v. Wilson Riles case that the plaintiffs had been placed in racially isolated, dead-end classes from which few students ever returned to the regular classroom. Peckham's ruling banned the use of IQ tests to place blacks in retarded classes. In 1986, he expanded the ban to the placement of black children in all special-education classes and programs.

Also in 1971, the Griggs v. Duke Power Co. decision by the U.S. Supreme Court struck a death blow at general aptitude tests administered by employers. In its ruling, the court placed the burden of proof on the employer to show that any test used to screen applicants was a “reasonable measure of job performance” and would not harm black job placement because of unequal test scores between racial groups. Under the Griggs standard, lawyers have overturned hundreds of employment tests used by civil service and private employers. Today, most employers are reluctant to use such tests, experts say.

Reaction to Jensen's Theories

Meanwhile, a simmering academic debate over the racial implications of IQ testing came to public attention after the Westinghouse Learning Corporation concluded in a 1968 report that Head Start hadn't produced long-term gains in the IQ scores of disadvantaged preschoolers. The next year, Arthur Jensen sparked a storm of controversy in the scholarly and popular press when he blamed the differences in black and white IQ scores on heredity.

Jensen has compared IQ scores of blacks and whites to the speed with which they react to simple stimuli, such as pushing a button when a light goes off. He finds a high correlation between IQ scores and these response times. [27] That research supports his view that the racial difference is predominantly genetic, he argues, because “there's some common factor in the brain that causes both reaction time and IQ tests.”

Public acceptance of IQ has waned considerably since the Jensen controversy. In 1973 and 1974, news stories reported that a British psychologist, Sir Cyril Burt, had faked data in twin studies showing that intelligence was inherited. The stories -- since disputed -- added to the cloud looming over IQ studies with a hereditarian, or genetic, bent.

Over the past decade, a new school of psychologists, led by Harvard's Howard Gardner, has spurned IQ tests as reflections of an outdated, narrow view of intelligence. In 1983, Gardner captured attention with his book Frames of Mind, which advanced his theory of multiple intelligences. He was the latest in a long line of thinkers who have argued since the 1900s that there are from four up to 300 forms of human intelligence.

But Gardner's views directly challenged the philosophical basis of IQ testing, first articulated by British psychologist Charles Spearman in 1904, that there is one predominant form of general intelligence. Adherents of that philosophy, led by Arthur Jensen, have been equally active in pursuing IQ's genetic and biological basis over the past decade. Ironically, the Jensen camp's research, though sophisticated in its use of modern medical technology, is reminiscent of Galton's simple sensory-response tests as the true measure of human intelligence.

go to top

*Test scores were obtained by comparing students' mentalage (derived from the number of items that should beanswered correctly by a normal child at that age) with theirchronological age. Today's Intelligence Quotient or IQderived from this measure. It is calculated by dividing anindividual's mental age by his real age and multiplying by100. An IQ of 100 (in which the mental and chronological ageare the same) is considered average. An IQ of 130 or higheris generally considered in the gifted range -- and from 140and up in the “genius” category. Children with IQs below 70-75 are considered mentally retarded by most school systems.

** The test had been translated into English in 1910 by HenryGoddard, director of research at a school for the retardedin Vineland, N.J. He was impressed by how well it accordedwith his intuitive judgments about who was “feeble-minded.”.


Chronology

19th Century
Charles Darwin's theory of evolution is used to explain differences in mental abilities and to rally support for eugenics, or selective breeding.

1859
Darwin publishes Origin of Species.

1869
Sir Francis Galton publishes a genealogical study of famous men, Hereditary Genius, supporting his view that intelligence is largely the result of inherited abilities.

1900s
The first intelligence tests provide the foundation for the IQ tests commonly used today.

1905
Alfred Binet and Theodore Simon publish their first intelligence scale for classifying the mentally retarded.

1908
Binet and Simon revise their scale to measure children's readiness for school.

1910
Henry Goddard translates the 1908 Binet-Simon scale into English.

1916
Lewis Terman publishes the Stanford Revision of the Binet-Simon scale, marking the start of large-scale individual intelligence testing in the U.S.

1917
Arthur Otis develops the first group intelligence test.

1918-1919
The Army administers intelligence tests to more than 1 million recruits.

1920s-1940s
As the racial and ethnic implications of widespread military testing become public, IQ tests come under attack.

1921
Army tests indicate lower scores for blacks and certain immigrants.

1922-1923
Walter Lippman challenges IQ tests in The New Republic.

1926
College Entrance Examination Board introduces the Scholastic Aptitude Test (SAT).

1941-1945
Nine million recruits take Army General Classification Test.

1950s-1960s
IQ testing in schools increases amid concern that Americans are falling behind technologically. Critics charge standardized testing blocks opportunities for black children.

1957
Launching of Soviet satellite Sputnik raises questions about U.S. education and the need for more testing.

1967
In Hobson v. Hansen, a judge rules that using aptitude tests to track students is racially biased.

1968
Westinghouse Learning Corporation reports that Head Start fails to produce long-term IQ gains for disadvantaged preschoolers.

1969
Psychologist Arthur Jensen suggests that intelligence is largely hereditary.

1970s-1980s
Lawsuits charging that IQ testing discriminates against minorities lead to diminished use of IQ tests in schools and workplace. Psychologist Howard Gardner proposes his theory of multiple intelligences.

1971
In Larry P. v. Wilson Riles, parents of seven black children sue California, claiming a culturally biased IQ test wrongly put their children in classes for the mentally retarded.

1980
Allan Nairn's book The Reign of ETS extends cultural-bias criticisms to the college SAT and the Law School Admissions Test (LSAT).

1983
Howard Gardner's book Frames of Mind describes his theory.

1986
U.S. District Judge Robert F. Peckham bans use of intelligence tests to refer black children in California to classes for the retarded.

1990s
Continuing concern about aptitude testing and racial bias leads to action.

1991
In response to charges of racial bias, the Labor Department discontinues the General Aptitude Battery Test.

Aug. 31, 1992
Judge Peckham rescinds his ban on IQ testing of blacks students.

July 1993
The National Research Council plans study of testing used by schools, employers and the military.

go to top


Current Situation

About three-quarters of the nation's school districts give IQ or aptitude tests two or three times during a child's school career, estimates Professor Carolyn M. Callahan of the Curry School of Education at the University of Virginia. In most cases, says Callahan, the test is intended to inform teachers whether students are working at the level indicated by their aptitude scores and to spot those needing remedial education.

In general, intelligence scores have the most influence on those at the high and low ends of the spectrum -- the gifted and the learning disabled. For the majority in the middle, group intelligence testing no longer has a key role in determining a child's tracking, says Harvard's Sheldon White. “What's happening now,” he says, “is IQ tests are not taken as seriously and are not used in the more enlightened schools to put kids in special classes without some kind of ‘contestation process'” from parents or teachers who may disagree with the placement. But teachers still use IQ scores as an objective back-up to their recommendations, according to White and other observers.

Intelligence tests play a major role in admission to gifted tracks. The majority of school districts used them to identify gifted children, according to a recent national survey. [28] Most use the scores in combination with other criteria, such as teacher recommendations.

“There's a trend toward placing much less emphasis on standardized intelligence testing” in such admissions decisions, Callahan says, considering it as only one part of a child's school profile. Some schools are experimenting with “performance testing,” which observes directly how children perform in such areas as science experiments, creative writing or art. (See story, p. 654.)

Since the 1980s, however, something of a backlash has emerged against the blanket condemnation of IQ tests. Last August, District Judge Robert F. Peckham rescinded his ban on testing black students in California for placement in special education. The ruling came in response to a suit brought by parents of black students who wanted their children tested as part of a special-education assessment. The parents' attorney, Mark Bredemeier, said his clients viewed the modern special education offered by California schools today as helpful to children with learning disabilities, not a dead-end track, as parents contended in the original 1979 Larry P. case.

However, IQ testing in California faces an uncertain future, since the state education department has announced its intention to develop an alternative form of assessment for all races. Some California school districts have already dropped IQ testing altogether. Meanwhile, Peckham's ruling is being appealed to the 9th Circuit by the original Larry P. plaintiffs, who initiated the ban on IQ testing.

End of “Race-Norming”

For decades, state and local employment offices have used the U.S. Department of Labor's General Aptitude Test Battery (GATB) to match up job applicants and employers. However, the test became the focus of controversy recently because some 35 state offices used “race- norming.” The practice allows black and Hispanic applicants to be referred to jobs on the basis of their test-score ranking within their own racial groups, not the entire population of job-seekers. Because blacks and Hispanics have lower average scores than whites, they may be referred to employers and hired before whites with higher scores. [29]

Critics charged that race-norming unfairly excluded better- qualified white workers from jobs, and Congress outlawed the practice in the 1991 Civil Rights Act. In December 1991, the Labor Department discontinued the controversial test “because of concerns over whether the test adequately serves all individuals, including minorities, veterans, those with disabilities and other workers.” [30] The department is redesigning the test.

However, defenders of race-norming saw it as a counterweight to the inequality of opportunity produced by the GATB, which, like IQ tests, results in black-white scoring gaps. The Lawyers Committee for Civil Rights Under Law, a leading critic of the test, maintained that it merely identified good test-takers, not necessarily those who would make the best employees. [31]

Such traditional-style aptitude tests, say theorists like Sternberg and Gardner, fail to measure the more creative thinking needed in the modern workplace. Their theories hold enormous appeal for educators interested in unearthing the talents of poor and minority schoolchildren, who often perform poorly on IQ tests and similar pencil-and-paper tests.

Putting Theories to Work

To identify children for its gifted track, the Charlotte- Mecklenburg, N.C., school system has started testing for Gardner's multiple intelligences as an alternative to IQ tests. In the 1992-1993 academic year, its first year of experimentation, 17 percent of second-graders identified as gifted were black, compared with about 10 percent under the old method in previous years.

“Once again, we seem to be in a crisis mode in this country because we believe we are not going to have a work force that can handle changes in technology and changes in economic competition,” says Michael Feuer of the National Research Council's Board on Testing and Assessment. The board, a scientific forum that advises employers, educators and the military, is expected to consider seriously the experimental-testing approaches inspired by the multiple-intelligence theorists.

But the new school of intelligence theory faces an uphill battle in convincing psychologists and school administrators that what it describes can be reliably measured. “It's another educational fad, but the evidence is soft,” says Case Western Reserve University psychologist Detterman. Detterman's own current research has a biological bent, relating intelligence to specific sets of genes.

According to Fortune's Seligman, “The stated premise of affirmative action is that minorities have subpar economic performance because of discrimination, and I think the IQ data are putting forward an alternative or perhaps a supplementary explanation. It isn't just discrimination.”

go to top


Pro/Con

IsQheestredictorfoberformance?

Linda S. Gottfredson
Professor of Educational Studies, University of Delaware.
From Intelligence Versus Training: Job Performance and Black-White Occupational Inequality, 1986.

A longstanding claim among social scientists has been that intelligence has little value for predicting differences in job performance. This claim has been laid to rest, however, by meta- analyses of the many personnel selection studies during the last half century. The evidence shows that intelligence predicts differences in performance in all jobs, and it predicts especially well in complex, high-level jobs. Intelligence is not the only correlate of job performance, of course, but it is the single most useful predictor.

Most of the studies included in those meta-analyses used supervisor ratings as the indicator of job performance, because such ratings are usually all that are available. Some people have objected that supervisor ratings are subject to considerable bias, and they have predicted that intelligence would have little relation to more objective measures of job performance. However, recent evidence indicates that intelligence predicts performance on objective work samples even better than it does supervisor ratings....

There are a variety of other arguments disputing the importance of IQ in the workplace, but I would like to focus here on the set of expectations that I refer to as the training hypothesis.... Intelligence has no direct impact on job performance [the theory goes], only an indirect one. Differences in job performance can be traced to differences in job skills and knowledge, and differences in skills and knowledge can, in turn, be traced to differences in education, training and experience....

A variant of this argument is the popular lay theory that job experience is a great equalizer. Differences in intelligence may lead to differences in performance among recent hires, it is conceded, but those differences wash out as workers become more experienced on the job....

The first problem with the training hypothesis is that its premise is implausible. It is not feasible to break the link between intelligence and performance in training. More intelligent people learn complex tasks and knowledge faster than do less intelligent people, and the more complex the task, the larger the absolute differences are in the amount of time individuals require to master the task....

At best, training produces mastery of (actually, minimum or average competence in) only some fraction of a job, and brighter people will master the remaining fraction faster or better than will less intelligent people.

Psychological Science
February 1993.
Conventional intelligence tests were originally designed by Alfred Binet and Theodore Simon to predict performance in school, and they provide fairly credible, although certainly incomplete, prediction of such performance, as well as of performance in job training. But their prediction of performance on the job is weaker....

Why is prediction of job performance by intelligence tests modest, or at least more modest than is prediction of academic performance? We believe ... that there are two important reasons. First, academic and practical problems have different characteristics.... As a result, tests of academic problem-solving ability result in lesser prediction of practical job-related performance than of academic or even job- learning performance.

We also believe that academic and practical intelligence are rather different in kind. In academic intelligence, the relevant knowledge is of content and rules, and is formal and out in the open. It is learned primarily by reading and listening, and it is highly valued in the schools. It is measured by conventional ability tests....

In practical intelligence, in contrast, the relevant knowledge is of norms, and the knowledge is informal and often tacit. It is knowledge about, rather than of, a discipline. It is learned primarily by observation and modeling, and it is devalued in most schools. It is probably best measured by simulations....

We believe the key to practical intelligence is what Michael Polanyi has called tacit knowledge, the practical know-how one needs for success on the job. Often it is not openly expressed or stated, and it is usually not taught directly....

Tacit knowledge increases, on average, with job experience, but is not a direct function of job experience. What matters most is not how much experience a person has, but how well the person utilizes the experience to acquire and use tacit knowledge. Tacit knowledge is not a fancy proxy for IQ. It almost never correlates significantly with IQ.... Tacit knowledge is not a proxy for measures of personality, cognitive style, or interpersonal orientation, either....

We are not condemning conventional theories or tests of intelligence, but in the prediction of job performance, we do believe that they need to be supplemented by broader ability measures.

go to top


Outlook

Responding to dissatisfaction with current tests used in schools and the workplace, educators and theorists are expressing growing interest in new forms of intelligence assessment. “In the educational community,” says Feuer, “there's a growing sense that these [traditional IQ] tests are picking up on a small fraction of the abilities we want young people to acquire in school and that they're used prematurely to label children as unable to learn.”

Because multiple-choice tests tend to measure “bit-by-bit knowledge,” he says, some experts fear they drive teaching in a similar direction. In the writing component of traditional intelligence tests, for example, students simply identify the components of good and bad writing but do not have to demonstrate that they can actually write. While this may be a good test of proofreading, it's not necessarily a test of writing skill, critics say.

And while aptitude tests tend to focus attention on those at the very high and low ends of the analytical spectrum, they do not give much information about the many students who score in the middle, nor do they give information about how to help them achieve. “We could be doing a better job for the 30 to 40 percent of the population not in the faster tracks,” says Feuer. “That's a lot of what the reform movement is about today -- elevating the achievement of kids in the middle track.”

“Performance testing” is the term for a new kind of assessment that would look at what kids can actually do and how they do it in a learning situation. It can entail essay writing instead of multiple- choice questions, observing a child conducting a scientific experiment or studying a student's creative work over time. As of 1991, according to the Office of Technology Assessment (OTA), 36 states were evaluating student writing by using direct writing samples, and 21 states had implemented other types of performance assessments. [32]

The Portfolio Approach

One of the most intriguing approaches uses multifaceted portfolios of children's work -- including drawings, creative writing and arithmetic problems -- to study how well a child grasps new concepts and uses them over time. In 1992, five states (Alaska, California, North Carolina, Rhode Island and Vermont) were using portfolios as mandatory, voluntary or experimental components of statewide educational-assessment programs. [33]

Although the concept is new to elementary schools, portfolios have been used in the college-level Advanced Placement (AP) examination for studio art for nearly 20 years. The assessment consists entirely of artwork, judged by a jury of art teachers who independently score the portfolio on a scale of 1 to 5. There are no essays or questions to answer.

At the Educational Testing Service, which administers the AP exam, Michael Ziki, executive director of training and technical assistance, says it would not be easy to transfer a portfolio approach to the SATs. “The cost would be tremendous,” he says, because people rather than machines would have to do the scoring. And there's the question of reliability. “Would the same two people looking at the portfolio give it the same score without a lot of training?” he asks. Ziki concedes that the SAT is limited to “verbal and quantitative skills. You can list 10,000 other things it does not measure.”

The questions Ziki raises about portfolio assessments are often asked about the new elementary school testing approaches based on Howard Gardner's multiple intelligence theory. In one exercise developed by Gardner's Project Spectrum, students are given the parts of an oil pump to put together. Sometimes they are handed gears and grinders to see what they will do with them. *

The primary challenge with these novel tests, says Caroline Callahan, is to train teachers to use them so that schools consider them as reliable as pencil-and-paper tests. “Standardized tests are easy,” she says. “You open up the book. The other stuff takes a lot of time.” She concedes that such innovative testing “won't go anywhere unless the research is done. It's going to take people a long time to convince schools this does yield more data than an intelligence test.”

Many of Gardner's followers believe that his approach will identify more talented minority and low-income children than traditional IQ tests. But there are doubts. For one thing, notes OTA, a test based on human judgment rather than machine scoring could “exacerbate differences between groups of test-takers from different backgrounds.” [34]

The president of the San Diego City Board of Education also worries that portfolios could penalize minority children. “It's all subjective,” Shirley Weber has noted. [35]While the evidence is both sparce and convicting, some studies report that the gaps between scores of blacks and whites are about the same on essay writing as on multiple-choice tests of reading comprehension.

In the long run, the new intelligence theories may have more impact on the nature of teaching itself, rather than how students are sorted, says Mindy Kornhaber, who is co-authoring a textbook on intelligence with Howard Gardner. “If you want engineers,” she says, “you can test for high math skills, but in theory we think you can also get people who have these spatial and mathematical skills without the great testing skills. You can also educate people better if you don't gear everyone's education around paper and pencil.”

While some critics see this approach as special pleading for the less bright, Kornhaber defends it as part of the great American tradition of giving people new opportunities “to demonstrate and build their strengths.”

Sarah Glazer is a free-lance writer in Washington, D.C., who reports on education, health and social-policy issues.

go to top

*Project Spectrum attempts to put Howard Gardner's theoryon multiple intelligences to practical use in the classroom.


Footnotes

[1] Arthur R. Jensen, “How Much Can We Boost IQ and Scholastic Achievement?” Harvard Educational Review, February 1969, No. 1, pp. 1-123.

[2] Mark Snyderman and Stanley Rothman, The IQ Controversy, the Media and Public Policy (1988), p. 58.

[3] See Gina Kolata, “Study Raises the Estimate of Inherited Intelligence,” The New York Times, Oct. 12, 1990, p. A22 and Thomas J. Bouchard Jr., et al., “Sources of Human Psychological Differences: The Minnesota Study of Twins Reared Apart,” Science, Oct. 12, 1990, pp. 223-250.

[4] For background, see “Head Start,” The CQ Researcher, April 9, 1993, pp. 289-312.

[5] Howard Gardner, The Unschooled Mind (1991), p. 12.

[6] Stephen J. Ceci, “How Much Does Schooling Influence General Intelligence and Its Cognitive Components? A Reassessment of the Evidence,” Developmental Psychology, fall 1991, pp. 703-721.

[7] Ibid., p. 703.

[8] Howard Gardner, Frames of Mind (1993), p. xv.

[9] Malcolm James Ree and James A. Earles, “Intelligence Is the Best Predictor of Job Performance,” Current Directions in Psychological Science, June 1992, pp. 86-89.

[10] Linda S. Gottfredson, “Intelligence Versus Training,” paper presented to the American Psychological Association, Aug. 24, 1986.

[11] Daniel Seligman, A Question of Intelligence (1992), p. ix.

[12] Office of Technology Assessment, Testing in American Schools: Asking the Right Questions, February 1992.

[13] This Sidney Harris cartoon was cited by Stanford University psychologist Robert Calfee to illustrate the important role that the cultural context plays in intelligence. Robert Calfee, “Paper, Pencil, Potential, and Performance,” Current Directions in Psychological Science, February 1993, pp. 6-7.

[14] Seligman, op. cit., p. 159.

[15] Ibid., p. 151.

[16] “ETS Finds Gender Bias on SAT -- Math,” FairTest EXAMINER, spring 1992, p. 4.

[17] See John U. Ogbu, Minority Education and Caste: The American System in Cross-Cultural Perspective (1978).

[18] Seligman, op. cit., p. 154.

[19] Charles R. Henderson Jr. and Stephen J. Ceci, “Is It Better to Be Born Rich or Smart?” in Keith R. Billingsley et al., eds., Scientific Excellence in Supercomputing, Vol. 2 (1992), pp. 706- 751.

[20] Quoted in Sheldon H. White, Notes on the Growth of Psychological Testing, Harvard University, unpublished paper.

[21] Robert S. Siegler, “The Other Alfred Binet,” Developmental Psychology, Vol. 28, no. 2 (1992), pp. 179-190.

[22] White, op. cit., p. 8.

[23] For background, see “Why Schools Still Have Tracking,” Editorial Research Reports, Dec. 28, 1990, pp. 745-760.

[24] Snyderman and Rothman, op. cit., p. 18, and Office of Technology Assessment, op. cit., p. 120.

[25] Office of Technology Assessment, op. cit., p. 125.

[26] Henderson and Ceci, op. cit., p. 735, and Christopher Jencks and James Crouse, “Aptitude vs. Achievement,” The Public Interest, spring 1982, p. 23.

[27] Arthur R. Jensen, “Understanding in Terms of Information Processing,” Educational Psychology Review, Vol. 4, No. 3, 1992, pp. 271-308, particularly pp. 299-300.

[28] Laurie B. Abeel, Carolyn M. Callahan and Scott L. Hunsaker, “The Use of Published Instruments in the Identification of Gifted Students” (unpublished paper).

[29] For background, see “Racial Quotas,” The CQ Researcher, May 17, 1991, p. 289.

[30] Department of Labor news releases, July 10, 1990, and Oct. 23, 1992.

[31] “Testing ‘Red Herring' Blocks Civil Rights Bill,” FairTest EXAMINER, summer 1991, pp. 16.

[32] Office of Technology Assessment, op. cit., p. 204.

[33] Ibid.

[34] Ibid.

[35] Ibid., p. 246.

go to top


Bibliography

Books

Fancher, Raymond E., The Intelligence Men: Makers of the IQ Controversy, W.W. Norton and Co., 1985. This is a good, clear explanation of the competing philosophies that gave rise to intelligence testing, presented in the form of mini- biographies of the major historical figures from Sir Francis Galton to Arthur Jensen.

Gardner, Howard, Frames of Mind: The Theory of Multiple Intelligences, Basic Books, 1993. In a new introduction to this 10th anniversary edition, Harvard University psychologist Gardner defends his theory of multiple intelligences in the face of conflicting biological evidence.

Gardner, Howard, The Unschooled Mind: How Children Think and How Schools Should Teach, Basic Books, 1991. Writing for parents and experts alike, Gardner describes how he would use his multiple-intelligences theory to reform teaching, encouraging hands-on, project-oriented activities.

Gould, Stephen Jay, The Mismeasure of Man, W.W. Norton and Co., 1981. This indictment of the IQ test and its historical connection to racist philosophies continues to be cited by both IQ critics and proponents.

Seligman, Daniel, A Question of Intelligence: The IQ Debate in America, Birch Lane Press, 1992. In this readable defense of IQ testing, Fortune columnist Seligman argues that a broad shift to testing in the work force would increase the nation's productivity by some $50 billion.

Mark and Stanley Rothman, The IQ Controversy, the Media and Public Policy, Transaction Books, 1988. The authors contrast press coverage of IQ testing -- which they consider unfairly negative -- to experts' views -- which are generally positive toward the test.

Articles

Aldhous, Peter, “The Promise and Pitfalls of Molecular Genetics,” Science, July 10, 1992, pp. 164-165. This article summarizes recent research linking IQ to genetics through studies of twins and molecular genetics.

Cohen, Deborah L., “New Study Links Lower IQ at Age 5 to Poverty,” Education Week, April 7, 1993, p. 4. A University of Michigan study finds that children who lived in “persistent poverty” for the first five years of their lives had IQs nine points lower than those who didn't experience poverty.

Kantrowitz, Barbara, “He's the Next Best Thing: A Student of Genius,” Newsweek, June 28, 1993, pp. 48-49. This profile of multiple-intelligences guru Howard Gardner gives a sense of his cultlike status among admirers.

Sternberg, Robert J., “Ability Tests, Measurements and Markets,” Journal of Educational Psychology, Vol. 84, No. 2, , pp. 134-140. Yale psychologist Sternberg discusses the defects in today's intelligence tests and explains why testing companies are reluctant to adopt new tests.

Wilson, James Q., “Uncommon Sense about the IQ Debate,” Fortune, Jan. 11, 1993, pp. 99-100. In this favorable review of Daniel Seligman's book (see above), University of California-Los Angeles professor Wilson asserts that Americans are uncomfortable about discussing IQ because it forces them to face facts about inherited differences in intelligence.

Reports and Studies

Assessment, Technology, Testing in American Schools: Asking the Right Questions, February 1992. OTA, a research arm of Congress, discusses the policy implications of new, experimental testing approaches in this comprehensive report on aptitude and achievement tests.

go to top


Next Step

Periodical Abstracts database

 

Bias in IQ Testing

Baker, James N., “Battling the IQ-Test Ban,” Newsweek, July 27, 1987, p. 53. A ruling made by U.S. District Judge Robert Peckham of California states that IQ tests are racially and culturally biased. Mary Amaya is fighting for the parents' right to request IQ tests regardless of race.

Helene; Woo, Junda; Yoshihashi, Pauline, “Legal Beat: Judge Lifts California Ban,” The Wall Street Journal, Sept. 3, 1992, p. B8. In a case involving families of nine black schoolchildren, a federal judge ruled in favor of the families and lifted a statewide ban on the use of standardized intelligence tests for black children in California special-education classes.

Doyle, Jim, “Ban on Black Students' IQ Tests Overturned,” San Francisco Chronicle, Sept. 2, 1992, p. A1. In a major civil rights ruling, a federal judge in San Francisco onSept. 1, 1992,repealed his 1986 order that banned the use of IQ tests on black students in California public schools as a means to identify those who are slow learners.

Vincent, Ken R., “Black/White IQ Differences: Does Age Make the Difference?” Journal of Clinical Psychology, March 1991, pp. 266-270. A study was conducted on IQ differences between black and white adults and children to determine the possibility of racial bias of IQ tests. The results show a lessening of racial IQ differences in children.

Genetics and IQ Testing

Cravens, Hamilton, “A Scientific Project Locked in Time: The Terman Genetic Studies of Genius, 1920s-1950s,” American Psychologist, February 1992, pp. 183-189. Lewis M. Terman is well-known in the history of U.S. psychology for the Stanford Revision of the Binet-Simon intelligence tests and the Genetic Studies of Genius project.

Davis, Richard, “Biological tests of intelligence as culture fair,” American Psychologist, June 1993, pp. 695-696. For decades, the controversy concerning the viability of IQ tests has raged unceasingly. It is argued that such biological tests may indeed be culturally fair.

Influences on IQ Scores

Patricia A.; Just, Marcel Adam; Shell, Peter, “What One Intelligence Test Measures: A Theoretical Account of the Processing in the Raven Progressive Matrices Test,” Psychological Review, July 1990, pp. 404-431. The cognitive processes in the Raven Progressive Matrices Test are analyzed in terms of which processes distinguish between higher- scoring and lower-scoring subjects and which processes are common to all subjects and all items on the test.

“Children's Higher IQ Is Linked to Breast Milk,” The Washington Post, Jan. 31, 1992, p. A4. A study of 300 children who were born prematurely suggests that children who were fed breast milk scored significantly higher on IQ tests than children who received formula.

“Education That Works,” Christian Science Monitor, July 6, 1990, p. 20. An editorial supports the conclusions of a new study of premature babies indicating that a comprehensive education program beginning just after birth can lead to higher scores on intelligence tests and fewer behavioral problems.

Henry, Tamara, “Planning Gives Girls an Intellectual Edge,” USA Today, Nov. 9, 1992, p. D1. A new intelligence test developed by university psychologists Jack Naglieri and J. P. Das shows that girls actually may be smarter than boys in early school years because of a natural ability to plan ahead.

Leon J. Kamin, “Blackness, deafness, IQ, and g,” Intelligence, January 1993, pp. 37-46. The deaf data employed by Jensen (1985) and Braden (1984,1989) have been defectively and inconsistently reported and can provide no support to the claim that black-white differences in IQ are not environmental in origin, the authors state.

Waite, Teresa L., “Breast Milk for Premature Babies Tied to Higher Intelligence Scores,” The New York Times, March 4, 1992, p. C14. According to a new study, children fed breast milk as premature infants scored significantly higher on intelligence tests than children also born prematurely who had not received their mothers' milk.

“Young Children with Poor Sight Score Higher in IQ Tests,” Guardian, Aug. 21, 1990, p. 6. The British Association was told recently that poor eyesight in young children is linked to intelligence, and the difference between shortsighted children and others could be five or 10 points on the IQ scale.

Intelligence Testing

Flynn, James R., “Massive IQ Gains in 14 Nations: What IQ Tests Really Measure,” Psychological Bulletin, March 1987, pp. 171-191. Data from 14 nations reveal IQ gains as high as 25 points in a single generation. It is suggested that IQ tests do not measure intelligence but rather a weak causal link to intelligence.

Gardner, Howard, “Developing the Spectrum of Human Intelligences,” Harvard Educational Review, May 1987, pp. 187-193. The theory that the standard IQ test is not a valid measure of intelligence is discussed. Research has explored the possibility that seven intelligences exist, and should be measured individually.

Levine, Art, “Getting Smart about IQ,” U.S. News & World Report, Nov. 23, 1987, pp. 53-55. Recent research is redefining what intelligence is, and how it is measured. The limited scope of the traditional IQ test, which only examines one type of intelligence, is countered by new, more comprehensive tests.

Shafran, Avi, “Keep intelligence in perspective; It's only one measure of worth,” Chicago Tribune, April 21, 1993, p. 27. Avi Shafran comments on Daniel Seligman's book “A Question of Intelligence,” which discusses the significance of intelligence testing. The book is threatening to upset the apple-cart of politically correct education theory, and while there might be some benefit, the very fact that the volume is controversial has some troubling implications.

Testing of Infants and Preschoolers

“Infant Intelligence Tests,” Chicago Defender, May 27, 1989, p. 22. Harold E. Charles criticizes the trend toward testing infants' intelligence to determine how well they are likely to perform in school.

Norris, Bill, “Hurried on to the exam treadmill at the age of 4,” Times Educational Supplement, Feb. 15, 1991, p. 15. Four-year-olds are competing for admission to private nursery schools in New York City. An entrance exam, called the ERB after the Educational Records Bureau, has sparked an intense debate about the validity of intelligence tests for very young children.

Uses of IQ Testing

Barnett, W. Steven, “Benefits of Compensatory Preschool Education,” Journal of Human Resources, spring 1992, pp. 279-312. The long-term effects of compensatory preschool education are examined to determine its ability to produce meaningful long-term improvements in educational and economic success. Results indicate that compensatory preschool education can produce long-term gains in school success through contributions to cognitive abilities not adequately measured by IQ tests, and this success is accompanied by improvements in economic outcomes.

Dumaine, Brian, “The New Art of Hiring Smart,” Fortune, Aug. 17, 1987, pp. 78-81. To compete in a global economy, American corporations are restructuring. As part of their efforts to cut costs, companies are using sophisticated new techniques for identifying talent, including job simulation exercises for blue-collar workers and a renewed interest in IQ tests.

Humphreys, Lloyd G., “Commentary: What both critics and users of ability tests need to know,” Psychological Science, September 1992, pp. 271-274. Intelligence testing has been heavily criticized in recent years on grounds of racial and ethnic discrimination. The validity of intelligence tests is discussed on two levels: generalized description of the items in intelligence tests and the correlates of the total scores on those items.

Kelman, Mark, “Concepts of Discrimination in General Ability Job Testing,” Harvard Law Review, April 1991, pp. 1158-1247. Federal civil rights laws prohibit employers from screening potential employees by administering general intelligence tests if the tests are not related to job performance and have a disparate impact on a protected group. The argument that law takes too narrow a view of discrimination to capture all of the harms caused by such tests is offered.

Williams, Dick, “Juries, Not IQ Tests, Must Decide Capital Punishment,” Atlanta Journal Constitution, Dec. 16, 1989, p. A17. Dick Williams comments on the Catch-22 situation created by the Georgia Supreme Court ruling in which a convicted killer's score on an intelligence test determines whether he will go to jail or to the electric chair.

go to top


Special Focus

Finding Gifted Students...

Group Cites Examples of Cultural Bias

Trying to Link Genetics and Intelligence

Assessing Students Artistic Growth

States That Use Performance Assessment

Assessing Students' Math Skills

The kindergartener showed limited aptitude for numbers when she used pencil and paper. But she dazzled her teachers with the “bus game.” Pushing a toy bus, the little girl in Montgomery County, Md., kept an accurate running count of imaginary riders as a teacher's story sent the bus whizzing from one stop to another to pick up and drop off passengers.

The bus game is actually a carefully designed alternative to the traditional method of identifying gifted students. Now being tested in several school systems, it aims to unearth intellectual talent among children who usually get poor test results, especially those from poor families or homes where little or no English is spoken.

The experiment represents a practical application of Harvard University psychologist Howard Gardner's theory that there are seven types of intelligence. In addition to the linguistic and mathematical skills measured by traditional IQ tests, Gardner lists five other intelligences: spatial, musical, kinesthetic, interpersonal (the ability to know and understand others) and intrapersonal (knowing and understanding oneself). The bus game was developed at Gardner's Project Spectrum, which seeks to apply his ideas to early childhood education.

At Montgomery Knolls elementary school, which has been testing Gardner's concept for three years under a U.S. Department of Education grant, many of the students are recent immigrants and receive free or reduced-cost lunches.

“The basic idea here is to identify and nurture all of the intelligences in populations traditionally at risk for school failure,” says Jean Barton, project coordinator for Montgomery County's multiple intelligences model program. “All of these children come with strengths. What they don't come with is culture and language.”

It's not clear how many schools have latched onto Gardner's multiple intelligences (MI) theory, launched in his 1983 book, Frames of Mind. Gardner research assistant Mindy Kornhaber estimates the number at anywhere from several dozen to a few hundred.

Although Project Spectrum advises selected schools and conducts its own seminars, there is no “test kit” a school can send for. Neither has Gardner developed the standardized examination or numerical scores typical of traditional aptitude tests.

“It's not easy and not prepackaged the way a lot of things are,” notes Patricia O'Connell Ross, director of the Department of Education's Jacob Javits Gifted and Talented Education Program. The program is funding Montgomery County and other school systems experimenting with Gardner's approach.

Ross was drawn to Gardner's theory because it may answer a fundamental criticism of gifted programs -- that they are primarily elite oases for the white middle class. According to a 1988 Department of Education survey of eighth-graders, almost half of all gifted and talented students come from the nation's top socioeconomic quarter.#

At the Charlotte-Mecklenburg, N.C., school system, Gardner-type assessments are uncovering far more minority children than traditional tests, according to Anne Udall, coordinator of gifted and talented programs. After using the new approach for the first time this year, 17 percent of second-graders identified as gifted were minority, compared with 10 percent in past years.

By observing how children assemble pumps, build toothpick bridges or design imaginary schools, educators in Charlotte-Mecklenburg are responding to Gardner's charge that traditional intelligence tests “rarely assess skill in assimilating new information or in solving new problems.” In Gardner's words, information-oriented aptitude tests are biased toward “crystallized” rather than “fluid” knowledge.## And some children may not be good at both.

After Charlotte-Mecklenburg used Gardner's approach to pick gifted students, “Teachers said some of [the] kids did really awful on the standardized test,” reports Udall.

Skeptics in the field of gifted education also question whether Gardner's theory, or something else, explains poor test performance among disadvantaged and minority children. Education Professor James Gallagher of the University of North Carolina-Chapel Hill agrees that “we have a lot of hidden talent.” But he suggests that the reason is deprivation at home, not neglect of one of seven intelligences.

Like many firm believers in IQ tests, Gallagher doubts that many children who score low in IQ will prove exceptional in the other areas. “If Gardner does believe it happens, then all he has to do is trot out the kids and say, ‘Here they are.'”

Yet so far, data is sparse in this relatively young experiment. “This is difficult work still in its infancy,” Gardner concedes, “but it is possible to assess intelligences in legitimate contexts. Our initial efforts indicate that children ages 4-5 may indeed perform well on the non-traditional intelligence tests even when they are not strong in language and logic.”

Gardner also disputes the view of IQ traditionalists that it takes a genius-level IQ to become a great artist, dancer or musician. “My own view is that high performance in specific domains requires a normal IQ but no more,” he says.

Even admirers of Gardner's work acknowledge that his concept of intelligence seems “murky” to many educators, in the words of the Jacob Javits program's Ross. “It's theory,” she says. “It isn't proved. You either believe it or you don't.”

# NELS Study: Gifted and Talented Education, Department of Education, 1988, p. 39.

## Howard Gardner, Frames of Mind (1993), p. 18.

FairTest, an advocacy group in Cambridge, Mass., says IQ testsdiscriminate against test-takers who don't fit a white, middle-class profile, such as students brought up in homes where English isn't spoken. FairTest considers the following questions from the Wechsler Intelligence Scale for Children-Revised (WISC-R) as examples of cultural bias often found in IQ tests. The WISC-R test is administered orally in a one-on-one situation, usually to diagnose learning problems. A newer version of the test, known as the WISC-III, is available, but WISC-R is still used by some schools. The San Antonio- based Psychological Corp., which publishes the WISC-R test and conducts its own bias-review experiments, contends that none of the questions below is biased.

Question 1

Q: “How tall is the average American man?”

A: “Any answer from 5'7” to 5'11“ is correct”.

FairTest's analysis: Children often answer with the height of their own fathers. Asians and Latin Americans are often shorter than typical American males.

Question 2

Q: “What is the thing to do when you cut your finger?”

A: 2-point response: “Put a Band Aid on it.”

1-point response: “Go to the doctor (hospital)....Get it stitched up.”

0-point response: “Suck blood.... Don't panic.”

FairTest's analysis: Minority children usually perform poorly on this item. A Baltimore, Md., sociologist found that many inner-city youths answered “Go to the hospital” because they thought that “cut” meant a big cut. When the children were told that “cut” meant a little cut, almost all then responded “Put a Band Aid on it.”

Question 3

Q: “What are you supposed to do if you find someone's wallet or pocketbook in a store?”

A: 2-point response: “Find out whose it is and return it....Give it to the store owner (store guard, policeman).”

1-point response: “Try to find the owner.”

O-point response: “Make believe you didn't see it.....”

FairTest's analysis: According to psychologist Robert Williams, who is black, “Given the context today of the negative emphasis on black crime, black children pulling wallets or snatching purses, it would be suicide for a child to say, ‘I would pick it up and try to find the owner.'... They would be accused of having snatched it.”

Source: FairTest, “IQ Examples,” unpublished paper.

Eerie stories abound. In 1979, Oskar, who was raised by a Nazi family in Czechoslovakia, was reunited with his identical twin, Jack, raised as a Jew in Trinidad. Both habitually flushed the toilet before and after using it, and both enjoyed deliberately sneezing to startle people in elevators.#

The two men participated in a study of more than 100 sets of identical twins and triplets. Since identical twins have the same genes, the study assumed that similarities between twins separated at an early age were due to genetics and that differences resulted from their dissimilar environments. The study team, led by University of Minnesota psychologist Thomas J. Bouchard, found a high correlation between the IQs of identical twins reared apart and concluded that the contribution of genes to IQ is 70 percent.##

But such findings can have disturbing implications. The prime minister of Singapore, for example, has cited the Minnesota research in defending eugenic policies that encouraged middle-class citizens to bear children and discouraged child-bearing by the poor. And liberals traditionally bristle at the suggestion that intelligence is inherited just like blue eyes -- and is just as impervious to change.*

In fact, critics of the Bouchard study point out, correlation does not equal causation; just because two things happen together a lot doesn't mean one causes the other. Take the example of hair length and gender posed by sociologist Christopher S. Jencks. We know that short hair is highly correlated with American males. Social scientists could say, therefore, that hair length is caused by having a Y chromosome (the chromosome that determines maleness). But a change in fashion tomorrow (in other words, an environmental effect) could wipe out the apparent biological connection.

Critics further caution that there is a high possibility of “coincidental similarity” in the case of two people born in the same country at the same time and raised in similar middle-class homes in Western societies. Several of the identical-twin pairs studied by Bouchard, including Oskar and Jack, had met before participating in the study, a typical weakness in twin studies.

“Since there is no practical method for separating the physical and social effects of genes,” Jencks maintains, estimates of the amount of influence genes have on intelligence and other traits must include both factors. Even Bouchard has acknowledged that the identical genetic makeup of the twins he studied “makes it probable that their effective environments are similar.” A fretful infant, for example, elicits a different response from a parent than one with a sunny disposition, Bouchard notes.

The extent to which adult expectations can affect a child's intellectual performance was illustrated in the famous Pygmalion experiment in 1968. At the beginning of the school year, researchers (who have been faulted on methodological and philosophical grounds) gave elementary school teachers a list of children who supposedly had high aptitude-test scores. In fact, the children were picked at random. When the school year ended, the experimental children showed significantly greater IQ-test gains than their classmates, presumably in response to their teachers' high expectations.* * #

Currently, some researchers are trying to link IQ to specific sets of genes, much as medical researchers have identified the gene for some hereditary illnesses. Robert Plomin, a psychologist at Pennsylvania State University, is expected later this year to report on his research into gene similarities among high-IQ children.

While Plomin's approach may be ideal for pinpointing diseases caused by a single gene, such as Huntington's disease, scientists think it unlikely there is a single gene for intelligence. Rather, IQ is believed to be influenced by tens or even hundreds of genes. It may well be that each of the genes involved subtly influences a person's IQ, together with environmental factors.

# John Horgan, “Eugenics Revisited,” Scientific American, June 1993, pp. 123-131.

## Thomas J. Bouchard, et al., “Sources of Human Psychological Differences: The Minnesota Study of Twins Reared Apart,” Science, Oct. 12, 1990.

* Horgan, op. cit., p. 126.

* * Christopher S. Jencks, Rethinking Social Policy (1992), p. 95.

* * # Robert Rosenthal and Lenore Jacobson, Pygmalion in the Classroom (1968).

Student portfolios chart the development of visual skills inPittsburgh's Arts PROPEL program. Eighth-grader Dennis Biggs did each portrait of a classmate in three minutes. The first is a contour drawing; the second uses all circular lines; the third uses lines drawn with a ruler. GRAPHICS: Drawings.

Source: “Testing in American Schools: Asking the Right Questions,” Office of Technology Assessment, 1992

In 1991, the latest year for which statistics are available, 36states evaluated students by examining their writings, portfolios of their work, or hands-on demonstrations. Most states used the tests in coordination with traditional multiple-choice tests. GRAPHICS: Map.

Source: “Testing in American Schools: Asking the Right Questions,” Office of Technology Assessment, 1992

A logic problem from the 12th grade California Assessment Program test assesses a student's ability to detect and explain faulty reasoning. Students' answers to such open-ended tests help teachers to gauge students' understanding of broad concepts.

Question: James knows that half of the students from his school are accepted at the public university nearby. Also, half are accepted at the local private college. James thinks that this adds up to 100 percent, so he will surely be accepted at one or the other institution. Explain why James may be wrong. If possible, diagram your explanation. GRAPHICS: Drawing.

Source: “Testing in American Schools: Asking the Right Questions,” Office of Technology Assessment, 1992