6.6 Intelligence: Theories, Measurement, Heritability, and Metacognition

Intelligence: Theories, Measurement, Heritability, and Metacognition

The content outline lists intelligence under Cognitive (Thinking) and again under Developmental (Learning, Intelligence), and the closely related planning and metacognition under Thinking. GRE items typically test theorists and their models, IQ scoring arithmetic, and the logic of heritability.

1. Early Measurement: Galton, Binet, Terman, and Wechsler

• Francis Galton (Hereditary Genius, 1869) believed intelligence was inherited and tried to measure it with reaction time and sensory acuity. These measures did not predict academic success.
• Alfred Binet and Théodore Simon (1905) built the first practical intelligence test to identify Paris schoolchildren who needed special instruction. Items were ordered by the age at which typical children passed them, yielding a mental age (MA). Binet saw intelligence as complex and modifiable, not a fixed single quantity.
• William Stern (1912) proposed the intelligence quotient, and Lewis Terman at Stanford adapted Binet's test as the Stanford-Binet (1916), using the ratio IQ:

$\text{IQ} = \frac{\text{Mental Age}}{\text{Chronological Age}} \times 100$

A 10-year-old with a mental age of 12 has a ratio IQ of 120. Ratio IQ breaks down for adults because mental age stops rising.

• David Wechsler (Wechsler-Bellevue, 1939) introduced the deviation IQ, which locates a person relative to same-age peers on a scale with mean 100 and SD 15 (Section 15.1). He also reported separate verbal and nonverbal (performance) scores. Current editions report index scores; the WAIS-5 (2024) and WISC-V (2014) each report five primary indexes: Verbal Comprehension, Visual Spatial, Fluid Reasoning, Working Memory, and Processing Speed.
• Group tests: In World War I, Robert Yerkes's committee created the Army Alpha (for literate recruits) and Army Beta (nonverbal, for recruits who were not literate in English).

2. Theories of Intelligence

3. Stability, Prediction, and Extremes

• Stability: IQ scores become increasingly stable with age. In the Scottish Mental Survey, scores at age 11 correlated about .63 with scores at age 77 (Ian Deary and colleagues, 2000). Infant tests such as the Bayley Scales predict later IQ poorly, but infant habituation and novelty-preference measures predict childhood IQ modestly.
• Predictive validity: IQ predicts school grades (correlations around .5), years of education, and job performance (Section 16.4).
• Giftedness: Terman's longitudinal study, begun in 1921, followed about 1,500 California children with IQs of roughly 135 and above. As adults they were, on average, healthy, well-adjusted, and successful, contradicting stereotypes that gifted children are frail or maladjusted.
• Intellectual disability: DSM-5-TR requires deficits in intellectual functioning (typically an IQ about two standard deviations below the mean, roughly 70 plus or minus 5 to allow for measurement error) and deficits in adaptive functioning (conceptual, social, and practical domains), with onset during the developmental period. Severity (mild to profound) is rated from adaptive functioning, not IQ alone. Common genetic causes include Down syndrome (trisomy 21) and fragile X syndrome.

4. Heritability, Environment, and Group Differences

• Heritability estimates: Twin and adoption studies place the heritability of IQ at roughly .50 overall. Heritability rises with age, from about .20 in infancy to about .40 in childhood and .60 or more in adulthood (the Wilson effect), probably because people increasingly select environments that fit their genetic tendencies (active gene-environment correlation, Section 1.4).
• Environment matters: Adoption into higher-SES homes raises IQ, schooling raises IQ, and Eric Turkheimer and colleagues (2003) found heritability near zero among impoverished U.S. children but high in affluent families.
• The Flynn effect: James Flynn documented that average raw scores on IQ tests rose by about 3 IQ points per decade across the twentieth century in many countries, with the largest gains on fluid-reasoning tests. Gene pools cannot change that quickly, so the gains reflect environmental changes such as nutrition, schooling, smaller families, and more abstract thinking demands. Some countries have shown a slowing or reversal since the 1990s.
• Within-group versus between-group differences: Richard Lewontin's seed analogy: plant genetically varied seeds in two pots, one with rich soil and one with poor soil. Within each pot, height differences are largely genetic, yet the average difference between pots is entirely environmental. A high heritability within groups therefore says nothing about the cause of a mean difference between groups.
• Test bias: Psychometricians distinguish mean score differences from bias. A test shows predictive bias if it predicts a criterion (such as grades) differently for different groups, and differential item functioning (DIF) if test-takers of equal ability from different groups have different probabilities of passing an item. The APA task force report Intelligence: Knowns and Unknowns (Ulric Neisser and colleagues, 1996) summarized this evidence. Situational factors such as stereotype threat (Section 8.3) can also depress scores.

5. Metacognition and Planning

John Flavell (1979) introduced metacognition: knowledge and regulation of one's own cognition. Thomas Nelson and Louis Narens (1990) proposed a two-level framework in which a meta-level monitors the object-level cognition and then exerts control over it.

Planning depends on the prefrontal cortex and is assessed with tasks such as the Tower of London (Section 6.4). Expert problem solvers show stronger metacognitive regulation: they monitor progress, notice impasses, and switch strategies sooner than novices.