The Story of Psychology (45 page)

How well children of any age did with each test was compared with how well they did on others; those tests which were too easy for children of a given age were shifted to an earlier place in the sequence, those which were too hard, to a later one. To balance the scale, additional tests were added at the lower and upper ends. The results of the testing were compared with teachers’ estimates of the same children’s intelligence by the Pearsonian correlation method; the overall correlation was .48, or moderately high, thereby validating the scale. The correlation would have been still higher had not teachers, in estimating childrens’ intelligence, sometimes failed to take into account that some of the children were either younger or older than most of their classmates.

The most valuable aspect of the revision was that the entire scale was far more thoroughly “standardized” than Binet-Simon or Goddard-Binet-Simon; that is, the scores were based on results achieved with a large standard sample of normal, retarded, and superior children and adults. On this basis, a child or adult who scored 100 was average; one who scored 130 or better was more intelligent than 99 percent of the population at large; and one who scored 70 or below was less intelligent
than 99 percent of the population. Terman classified the grades of intelligence as follows:

Terman, a mild-mannered and kindly man, voiced benign hopes for the use of the new scale:

If the Stanford-Binet did not exactly make those sentiments a reality, neither did it, fortunately, make a reality of Terman’s vision of its use in eugenics:

The Stanford-Binet, published in 1916, swiftly became the standard test for measuring intelligence and remained so for over two decades. It was soon being used in a number of schools, preschools, colleges, and institutions for the feeble-minded. But its influence was both broader and more profound than that; the Stanford-Binet scale (and, later, its 1937 revision) became the standard for virtually all IQ tests that followed it. What Binet, Simon, and Terman took to be the attributes making up intelligence became the model for nearly all later intelligence
tests; these components included memory, language comprehension, size of vocabulary, eye-hand coordination, knowledge of familiar things, judgment, likenesses and differences, arithmetical reasoning, ability to detect absurdities, speed and richness of association of ideas, and several others.
⁴⁹

A subsequent test using Stanford-Binet components revolutionized the field of intelligence testing.

All versions of the Binet scale—eventually there were dozens—have to be given by a psychologist or trained technician to one person at a time. But group testing, in which subjects read questions to themselves and check off multiple-choice answers or make appropriate marks on the form, would be far quicker, simpler, and very much less expensive.

This breakthrough in mental measurement came about as a result of the entry of the United States into the First World War. Within two weeks of President Woodrow Wilson’s signing of the declaration of war, on April 6, 1917, the American Psychological Association appointed a committee to see what role psychology could play in the war effort. The committee reported that the most useful contribution of the profession would be the development of psychological examinations that could be quickly given to large numbers of military personnel so as to eliminate the mentally incompetent, classify individuals according to their abilities, and select the most competent for special training and responsible positions.

A group of psychologists—among them Terman, Goddard, and Robert Yerkes, a Harvard professor—met at Vineland and began planning the tests. In August, Yerkes was commissioned a major in the Army and was ordered to carry out the plans. He assembled a staff of forty psychologists who, in two months, produced the Army Alpha, a written test of intelligence, and the Army Beta, a pictorial test for the 40 percent of inductees who were functionally illiterate (the instructions for Beta were read aloud by an assistant). The widely used Alpha looks, from today’s perspective, like a curious mixture of scientific information, folk wisdom, and morality, as can be seen by these questions:

1. If plants are dying for lack of rain, you should

Yerkes’ team began giving the tests in four camps, but within weeks the surgeon general decided to extend the program to the entire Army; by the time the war ended, in November 1918, more than 1.7 million men had taken the tests and some three hundred psychologists under Yerkes had graded each man and suggested a suitable military assignment for him.
⁵⁰
Although Yerkes’ psychological corps met resistance and noncompliance from professional officers, the tests resulted in the discharge of about eight thousand men as unfit and the assignment of about ten thousand of a low level of intelligence to labor battalions and similar services. Of greater importance, the Alpha was a factor in the selection of two thirds of the 200,000 men who became commissioned officers during the war.
⁵¹

The Army Testing Program, however, had far greater impact outside the military than within it. It made America more conscious than ever of the practical applications of psychology, specifically those derived from mental measurement. (James McKeen Cattell said that the war had put psychology “on the map,” and G. Stanley Hall that it had given psychology an invaluable redirection toward the practical rather than the “pure.”)

The Alpha, in particular, led to an explosive expansion of intelligence testing, which rapidly became a multimillion-dollar industry. Within a few years of the war’s end, a number of Alpha-type paper-and-pencil intelligence tests were being marketed to school administrators throughout the country. One of the most successful, appearing in 1923, was put together by Terman, Yerkes, and three co-workers, under the auspices of the National Research Council, which advertised it as “the direct result of the application of the army testing methods to school needs.” By the end of the decade it had been given to seven million American school-children.
⁵²
Another major success was the Scholastic Aptitude Test, developed by Carl C. Brigham, a colleague of Yerkes’, from Army models. Testing became prevalent in schools, colleges, the military services, institutions of all sorts, and various segments of industry.

The widespread use of intelligence testing was given further impetus by statistical evidence that the tests measured not just a series of separate mental aptitudes but an innate core of mental ability or “general intelligence.” Charles Spearman, an English psychologist and statistician, had shown that many mental abilities are correlated. (A person who does well in vocabulary, for instance, is likely to do well in arithmetic and other subtests.) He took this to indicate that an inherent general intelligence, which he called
g
, underlay all the specific abilities. Even if intelligence tests relied in part on learning, the correlations implied the existence of an innate ability to learn.

This provided additional justification for intelligence testing in the schools, which by the 1930s, both in the United States and in Great Britain, were classifying pupils early in the educational process and assigning them to broad programs preparatory for higher education or to narrow “vocational” or “technical” programs readying them for blue-collar jobs. In America this was called “tracking” and in Great Britain “streaming.”
⁵³

The growth of testing was not limited to the measurement of intelligence. During the 1920s and 1930s many other scales were developed to measure musical, mechanical, figural, verbal, mathematical, and other abilities, and a number of vocational aptitudes. Even though intelligence testing itself came under attack as early as the 1920s, Binet’s approach to mental testing had opened up a vast new area of psychological research, and the Army Alpha had transformed his cumbersome and costly procedure into one that was easy and inexpensive enough to be the psychological equivalent of the assembly line.

Intelligence testing did not long maintain its unquestioned status. From 1921 on, when Yerkes edited a massive report of the findings of the Army Testing Program, intelligence testing came under attack by various advocates and spokespersons of the underprivileged, who claimed that it measured not innate intelligence but acquired knowledge and cultural values and therefore was biased in favor of the dominant white middle class and against the lower classes and immigrants.

The Alpha, they charged, measured not native intelligence but the kinds of knowledge possessed by men with schooling and a degree of
sophistication. Here, for instance, is a typical example of the culturally biased kind of question:

The Beta was similarly far from impartial: in it, illiterates had to complete certain pictures such as a face without a mouth—fair enough—but others, such as a lightbulb without a filament or a tennis court without a net, made many lower-class men and immigrants seem stupid.
⁵⁴

The same criticism was made, and rightly so, as far as the Stanford-Binet scale was concerned. Many or most of the items in it measure a combination of inherent ability and acquired information or skills, but a person who has had little chance to acquire the information or skills will do poorly with the questions, no matter what his or her inherent mental power.

At the twelve-year-old level, for instance, the Stanford-Binet asked for definitions of “charity” and “justice.” If a Mexican-American child from a rural southwestern shanty town gave inadequate answers, did that indicate a lack of innate intelligence or the child’s failure to learn the meanings of those concepts in white, middle-class America? Again, at the eight-year-old level Stanford-Binet asked, “What’s the thing for you to do when you have broken something which belongs to someone else?” If the eight-year-old lived in a city slum where children struggled to survive, was his or her answer a gauge of innate intelligence or of the mores and folkways of the slum subculture?

Binet had left moot the extent to which mental development, measured by his scale, was due to heredity and to experience. But the tenor of Terman’s comments in
The Measurement of Intelligence
(the Stanford-Binet instruction manual), despite the disclaimer quoted above, was that intelligence is largely hereditary and that poor scores reveal mental deficiency—which, he said, was a genetic and a racial trait:

In 1922, the respected columnist and pundit Walter Lippmann launched a critical attack in
The New Republic
on Terman, Yerkes, and others who claimed that intelligence testing measured innate mental ability. Lippmann sounded the theme, repeated from that time until now, that such testing stamped a permanent label of inferiority on children, especially the underprivileged, and served the purposes of the prejudiced and the powerful.
⁵⁶

He and others who shared his views had even stronger grounds for objecting to the Army Alpha and Beta than to the Stanford-Binet, and for disputing Yerkes’ claim that tests modeled on the Alpha “measure native intellectual ability.” The answers to many Alpha questions clearly required learned information rather that intelligence, as Stephen Jay Gould later made plain in his polemical study,
The Mismeasure of Man
, in which he cited these examples:
⁵⁷