Binet's Brainstorm Wrestling with the problem, Binet had a great insight: In the classroom, the responses of “dull” children resembled those of ordinary children of younger ages. Bright children, in contrast, responded like children of older ages. The thing to measure, then, was a child's mental age (MA), or level of intellectual development relative to that of other children. Then instruction could be tailored to the child's capabilities.

The test devised by Binet and his colleague, Théodore Simon, measured memory, vocabulary, and perceptual discrimination. Items ranged from those that most young children could do easily to those that only older children could handle, as determined by the testing of large numbers of children. A scoring system developed later by others used a formula in which a child's mental age was divided by the child's actual age to yield an intelligence quotient (IQ) (a quotient is the result of division). With this formula, all average children, regardless of their age, would have an IQ of 100 because their mental age and their actual age would be the same. But a child of 8 who performed like the average 10-year-old would have a mental age of 10 and an IQ of 125 (10 divided by 8, times 100). For a brief history of the development of intelligence testing, watch the video Intelligence Testing, Then and Now 1.

Unfortunately, this method of computing IQ had serious flaws. At one age, scores might cluster tightly around the average, whereas at another age they might be more dispersed. As a result, the score necessary to be in the top 10 or 20 or 30 percent of your age group varied, depending on your age. Also, the IQ formula did not make sense for adults; a 50-year-old who scores like a 30-year-old does not have low intelligence! Today, therefore, intelligence tests are scored differently. The mean (average) is usually set arbitrarily at 100; tests are constructed so that about two-thirds of all people score between 85 and 115; and individual scores are computed from tables based on established norms. These scores are still informally referred to as IQs, and they still reflect how a person compares with other people, either children of a particular age or adults in general. At all ages, the distribution of scores approximates a normal (bell-shaped) curve, with scores near the mean more common than high or low scores (see Figure9.8).

Figure 9.8

The Expected Distribution of IQ Scores

The IQ Test Comes to America In the United States, Stanford psychologist Lewis Terman revised Binet's test and established norms for American children. His version, the Stanford–Binet Intelligence Scales, was first published in 1916, and has been updated several times since. It can be given to children as young as age 2 or adults as old as age 85. The test asks a person to perform a variety of tasks—to fill in missing words in sentences, answer questions requiring general knowledge, predict how a folded paper will look when unfolded, measure a quantity of water using two containers of different sizes, and distinguish between concepts that are similar but not exactly the same (such as, say, vigor and energy). The older the test taker is, the more the test requires in the way of verbal comprehension and fluency, spatial ability, and reasoning.

Two decades later, David Wechsler designed another test expressly for adults, which became the Wechsler Adult Intelligence Scale (WAIS); it was followed by the Wechsler Intelligence Scale for Children (WISC). These tests, too, have been updated several times. They produce a general IQ score and also separate scores for verbal comprehension, perceptual reasoning, processing speed, and working memory (the ability to hold information in mind so that it can be used for a task). Items measure a range of abilities, including vocabulary, arithmetic abilities, the ability to recognize similarities (e.g., “How are books and movies alike?”), general knowledge and comprehension (e.g., “Who was Thomas Jefferson?” “Why do people who want a divorce have to go to court?”), and nonverbal skills, such as the ability to re-create a block design within a specified time limit or to identify a part missing from a picture. (See Figure9.9.) More information on these types of tests is provided in the video Intelligence Testing, Then and Now 2.

Binet had emphasized that his test merely sampled intelligence and did not measure everything covered by that term. A test score, he said, could be useful, along with other information, for predicting school performance, but it should not be confused with intelligence itself. The tests were designed to be given individually, so that the test giver could tell when a child was ill or nervous, had poor vision, or was unmotivated. The purpose was to identify children with learning problems, not to rank all children. But when intelligence testing was brought from France to the United States, its original purpose got lost at sea. IQ tests became widely used not to bring slow learners up to the average, but to categorize people in school and in the armed services according to their presumed “natural ability.” The testers overlooked the fact that in America, with its many ethnic groups, people did not all share the same background and experience (Gould, 1996).