Measuring Intelligence

Measuring Intelligence

The goal of most intelligence tests is to measure “g,” the general intelligence factor. Good intelligence tests are reliable (consistent over time) and valid (they actually measure intelligence rather than something else). Because intelligence is highly individualized, psychologists have invested substantial effort in creating and improving these measures.

The First Intelligence Test

From 1904-1905, French psychologists Alfred Binet (1857–1914) and Théodore Simon (1872–1961) worked for the French government to develop a measure to identify children who would be unsuccessful in the regular school curriculum so they could receive specialized education (Aiken, 1994). Binet and Simon developed what most psychologists today regard as the first intelligence test, the Binet-Simon test. It consisted of a wide variety of questions that included the ability to name objects, define words, draw pictures, complete sentences, compare items, and construct sentences.

Binet and Simon believed that the questions they asked the children assessed the basic abilities to understand, reason, and make judgments (Binet & Simon, 1915; Siegler, 1992). The correlations among these different types of measures were all positive; students who got one item correct were more likely to get other items correct as well, even though the questions themselves were very different from each other.

Other Intelligence Tests

Soon after Binet and Simon introduced their test, the American psychologist Lewis Terman at Stanford University (1877–1956) developed an American version of Binet’s test known as the Stanford-Binet intelligence test. The Stanford-Binet measures general intelligence through a variety of tasks, including vocabulary, memory for pictures, naming of familiar objects, repeating sentences, and following commands.

The Wechsler Adult lntelligence Scale (WAIS) is the most widely used intelligence test for adults, but there are also versions for children and adolescents (Watkins et al., 1995). The WAIS consists of different tasks to assess intelligence, including working memory, arithmetic ability, spatial ability, and general knowledge about the world. The WAIS-IV, the current version of the test, yields scores on four domains: verbal, perceptual, working memory, and processing speed. It is correlated highly with other IQ tests, criteria for academic and life success, and measures of everyday function for people with intellectual disabilities.

Measuring Infant Intelligence

Intelligence tests for infants originally measured the development of motor, sensory, and language skills (Brooks-Gunn & Weinraub, 1983). While these tests can help determine if a child is neurologically impaired or developmentally delayed, they are not highly correlated with cognitive abilities later in life (Slater, 1997). However, information processing tests that measure how well a infants acquire, remember, and retrieve information are correlated with later intelligence (Fagan et al., 2007).

One common information processing test measures visual recognition memory, or if an infant can distinguish between what is new or familiar to them (Fagan et al., 2007). This is often done through habituation tasks that measure attention and looking time. A researcher may show an infant an image of a car for 30 seconds, put it away, and then show 2 images: the original picture of the car and a new one of a tree. The researcher then records how long the infant looked at each image. Infants will usually spend about 60% of their time looking at the new image (Fagan et al., 2007). Habituation in infancy is correlated with later intelligence (Fagan et al., 2007).

Other techniques for measuring information processing test cross-modal transfer. Cross-modal transfer tests measure if an infant recognizes an object using a different sense than they originally did (American Psychological Association, 2018a). For example, the infant might be handed a textured ball to play with, but a screen prevents them from seeing it. The textured ball is taken away and the infant is shown images of both a smooth and textured ball. Which one will they spend more time looking at? Studies show that infants will look longer at the unfamiliar object, like the smooth ball, than they will at the familiar object, such as the textured ball (Gottfried et al., 1977; Rose et al., 2009). Previous studies show that infant cross-modal ability is positively correlated with intelligence at 11 years of age (Rose et al., 1998).

Performance on information processing tasks as an infant can predict an adolescent’s general intelligence, academic achievement, and specific cognitive abilities like perceptual reasoning and reading ability (Bornstein et al., 2013; Fagan et al., 2007; Rose et al., 1992).