Traditional Models of Intelligence
development than previously thought. We will therefore begin the next section with a discus- sion of how psychologists have traditionally defined intelligence, and then explore more recent, alternative models.
Section Review Explain how twin and adoption studies inform us about the relative influences of biological and environmental influences.
9.2 Traditional Models of Intelligence Traditional models of intelligence depend on comparing people with one another to deter- mine levels of intelligence. For better or worse, the history of intelligence is linked with intel- ligence testing (Comer & Gould, 2012). In some ways, we can even define intelligence as that which an intelligence test measures. Here we look at how psychometrics, the field of study that designs and uses standardized quantitative tests to measure psychological traits, has been used in the study of individual intelligence.
Alfred Binet and Mental Age At the beginning of the 1900s, psychologist Alfred Binet (1857–1911) was commissioned by the French Ministry of Education to identify students who would be predicted to need spe- cial help in school. As a result, Binet developed the first usable intelligence test and laid the foundation for intelligence testing that is still done today. He used the term mental age (MA), to denote the average cognitive abilities of a child at any particular age. Regardless of chrono- logical age, if a child could answer questions that were typical of a 6-year-old, but could not answer most questions that were typical of older children, that child was assigned a mental age of 6. Although Binet stressed the limitations of the test, interest in it quickly spread.
Soon thereafter, William Stern (1914) built on Binet’s idea and devised the intelligence quo- tient (IQ). To calculate an IQ, mental age is divided by chronological age (CA) and then mul- tiplied by 100. So if MA and CA are the same, then IQ is 100, the expected average (see Table 9.1). If MA is greater than CA, then IQ is greater than 100; if the reverse is true, then IQ is less than 100. For example, if a 10-year-old has the MA of a 12-year-old, IQ would be 12/10 × 100 = 120.
Table 9.1: Example of IQ calculations
(a) (b) (c)
Mental age (MA) 29 12 15
Chronological age (CA) 29 10 17
Formula (29/29) × 100 = 100 (12/10) × 100 = 120 (15/17) × 100 = 88
mos82599_09_c09_297-326.indd 302 2/11/16 8:37 AM
© 2016 Bridgepoint Education, Inc. All rights reserved. Not for resale or redistribution.
303
Section 9.2 Traditional Models of Intelligence
The Stanford-Binet Intelligence Test With the groundwork in place, the use of IQ soon became widespread. Stanford University professor Lewis Terman adopted Binet’s work and standardized the test for use with children in the United States. Published in the early 1900s, this became the well-known Stanford- Binet Intelligence Test (Terman & Merrill, 1960). Still widely used, the Stanford-Binet is currently in its 5th edition and provides norms for ages 2 to 85 (Roid, 2003). When it was originally developed, the Stanford-Binet provided a global measure of intelligence using IQ as a standard.
The idea of a single variable to describe intelligence was also consistent with the work of Charles Spearman (1863–1945). A psychologist who specialized in statistics, Spearman the- orized that intelligence could be quantified as g (general intelligence) and derived from statistical operations (Spearman, 1923). Spearman agreed that intelligence included many different kinds of mental activities; he simply felt that they could be synthesized mathemati- cally into one general factor. At the time, the Stanford-Binet was the standard that allowed comparisons between individuals on this one theoretical factor.
The Wechsler Intelligence Scales David Wechsler, who was trained to use the Stanford-Binet test, recognized the limitations of a test developed for children when he evaluated older, U.S. Army recruits during the late 1930s (Kaufman & Lichtenberger, 2006). His work eventually led to the Wechsler Adult Intelligence Scale, now in its fourth edition (WAIS-IV). The Wechsler Scales now include the WAIS-IV, the widely used Wecshler Intelligence Scale for Children (WISC-IV), as well as the Wechsler Preschool and Primary Scale of Intelligence (WPPSI-III).
Rather than one IQ number, newer versions of both the Stanford-Binet and the Wechsler Scales have subscales, which address strengths and weaknesses in several areas (scales) like abstract reasoning, attention, processing speed, and factual knowledge. This change in focus from a mathematical quotient that singularly describes intellect has been an important advancement in the assessment as well as the definition of intelligence. Today, depending on strengths and weaknesses on different scales, clusters of scores can be used to analyze and diagnose specific kinds of learning patterns. Certain profiles on the scales may indicate gifted- ness, a learning disability, or specific attention problems that may impact learning. They also can sometimes correlate with mental problems, which we will discuss more in section 9.3.
Another important advancement in psychometrics has been the development of the devia- tion IQ. Instead of mental age, the deviation IQ is based on standard deviation, a statistical measure that tells us how much a particular score deviates from the average. A significantly high or low deviation from 100 suggests someone who is significantly more or less intelligent than average. Scores are placed around a normal distribution (bell curve), where 100 is the average. By definition, the middle 68% of scores lie between an IQ of 85 and 115 (84 and 116 on the Stanford-Binet). (Because one standard deviation on the WISC-IV is 15 and one standard deviation on the SB5 is 16, a score of 115 on the WISC-IV is the equivalent of 116 on the SB5; 130 on the WISC-IV is the equivalent of 132 on the SB5, and so forth.) As Figure 9.2 indicates, an IQ of 130 represents a score higher than 97.7% of the population, and an IQ of 70 represents a score that is lower than 97.7% of the population; both scores deviate from the average by the same amount.