Intelligence

Four influential theoretical positions have had a major impact on empirical research on intelligence and age. The earliest conceptualization came from Sir Charles Spearman's work in 1904. Spearman suggested that a general dimension of intelligence (g) underlied all purposeful intellectual pursuits. All other components of such pursuits were viewed as task or item specific (s). This view underlies the family of assessment devices that were developed at the beginning of the twentieth century; particularly the work of Alfred Binet and Henri Simon in France. Thinking of a single, general form of intelligence may be appropriate for childhood, when measurement of intellectual competence is used primarily to predict scholastic performance. However, a singular concept is not useful beyond adolescence, because there is no single criterion outcome to be predicted in adults. Also, there is convincing empirical evidence to support the existence of multiple dimensions of intelligence that display a different life-course pattern.

The notion of a single dimension of intelligence became popular during World War I, when Robert Yerkes constructed the Army Alpha intelligence test for purposes of classifying the large number of inductees according to their ability level. Because of the predominantly verbal aspects of this test, it was soon supplemented by performance measures suitable for illiterate or low-literate inductees. Assessing a single dimension of intelligence also widely influenced educational testing. It was around this time that Lewis Terman, a psychologist working at Stanford University, adapted the work of Binet and Simon for use in American schools and introduced the Stanford-Binet test, which dominated educational testing for many decades. Terman was also responsible for the introduction of the IQ concept. He argued that one could compute an index (the intelligence quotient, or IQ) that represents the ratio of a person's mental age (as measured by the Stanford-Binet test) divided by the person's chronological age. An IQ value of 100 was assigned to be equivalent to the average performance of a person at a given age. The IQ range from 90 to 110 represented the middle 50 percent of the normal population. Because there is no linear relation between mental and chronological age past adolescence, however, the Stanford-Binet approach did not work well with adults (see below).

An early influential multidimensional theory of intelligence was Edward. L. Thorndike's view that different dimensions of intelligence would display similar levels of performance within individuals. Thorndike also suggested that all categories of intelligence possessed three attributes: power, speed, and magnitude (see Thorndike & Woodworth, 1901). This approach is exemplified by the work of David Wechsler. The Wechsler Adult Intelligence Scale (WAIS) consists of eleven distinct subscales of intelligence, derived from clinical observation and earlier mental tests, combined into two broad dimensions: verbal intelligence and performance (nonverbalmanipulative) intelligence. These dimensions can then be combined to form a total global IQ. The global IQ (comparable to the Stanford-Binet IQ) is statistically adjusted to have a mean of 100 for any normative age group. The range of the middle 50 percent of the population is also set to range from 90 to 110. The Wechsler scales were used in the clinical assessment of adults with psychopathologies, and some of the subtests are still used by neuropsychologists to help diagnose the presence of a dementia.

The Wechsler verbal and performance scales are highly reliable in older persons, but measurable differences between the two are often used as a rough estimate of age decline, a use that has not proven to be very reliable. A more significant limitation of the test for research on intellectual aging, however, has been the fact that the factor structure of the scales does not remain invariant across age. As a consequence, most recent studies of intellectual aging in community-dwelling populations have utilized some combination of the primary mental abilities.

Factorially simpler multiple dimensions of intelligence were identified in the work of Louis Leon Thurstone during the 1930s, which was expanded upon by J. P. Guilford in 1967. (The primary mental abilities described by Thurstone and Guilford have also formed the basis for this author's own work; see Schaie, 1996b.) Major intellectual abilities that account for much of the observed individual differences in intelligence include verbal meaning (recognition vocabulary), inductive reasoning (the ability to identify rules and principles), spatial orientation (rotation of objects in two- or three-dimensional space), number skills (facility with arithmetic skills), word fluency (recall vocabulary), and perceptual speed. Further analyses of the primary intellectual abilities have identified several higher-order dimensions, including those of fluid intelligence (applied to novel tasks) and crystallized intelligence (applied to acculturated information).

The introduction of Piagetian thought into American psychology led some investigators to consider the application of Piagetian methods to adult development. However, Jean Piaget's (1896–1980) original work assumed that intellectual development was complete once the stage of formal operations had been reached during young adulthood. Hence, this approach has contributed only sparsely to empirical work on adult intelligence.

There are also discernable secular trends that cut across theoretical positions on different aspects of adult intelligence. Diana Woodruff-Pak has identified four stages: (1) until the mid-1950s, concerns were predominantly with identifying steep and apparently inevitable age-related decline; (2) the late 1950s through the mid-1960s saw the discovery that there was stability as well as decline; (3) beginning with the mid-1970s, external social and experiential effects that influenced cohort differences in ability levels were identified; and (4) in the 1980s and 1990s the field has been dominated by attempts to alter experience and manipulate age differences. Successful demonstrations of the modifiability of intellectual performance has led researchers to expand definitions of intelligence and to explore new methods of measurement.