Class 28: Intelligence: Definitions and Notes on Measuring IQ

What does it mean to say that someone is "intelligent"? What are the ways in which intelligence is demonstrated? Can "intelligence" be measured? Where does intelligence come from: the environment? heredity? All of these are the sorts of questions that psychologists have been looking at for over a century. The topic of "Intelligence" could easily require multiple classes to study and understand the findings of all those psychologists. But, for this Introductory class, I am only going to highlight a couple of the major notions associated with the notion of "intelligence"

Some Definitions and Models of Intelligence

There are multiple ways in which psychologists have defined or constructed models of how to understand human intelligence,

Charles Spearman (1863-1945) & the Two-Factor Theory

Initially the English psychologist, Spearman argued that “intelligent behavior is generated by a single, unitary quality within the human mind or brain. Spearman derived this theoretical entity, called the general factor, or simply g, through a new statistical technique that analyzed the correlations among a set of variables. This technique, called factor analysis, demonstrated that scores on all mental tests are positively correlated; this oﬀered compelling evidence that all intelligent behavior is derived from one metaphorical pool of mental energy.” (Plucker & Esping, 2014)

Eventually, Spearman agreed that there were a limited number of speciﬁc abilities which varied from individual to individual that were labelled ’s’

In 1943 David Wechsler, the author of the most widely used set of tests of intelligence in the United States, proposed in a famous definition that

“Intelligence is the aggregate or global capacity of the individual to act purposefully, to think rationally, and to deal eﬀectively with his/her environment” (emphasis added).

Anne Anastasi (1992), one of my own teachers in graduate school who studied intelligence testing for decades proposed that

“Intelligence is not a single, unitary ability, but rather composite of several functions. The term denotes that combination of abilities required for survival and advancement within a particular culture.” (emphasis added)

Consider what an intelligence test might be like which uses content that is familiar in one cultural context, but not in another. There have been multiple research projects which have compared the performance of native and non-native adults on varying kinds of tests. For example, the indigenous people of Australia, the Aborigines, have been compared with Australians of European ancestry. When tested on Western-content intelligence tests, the Aboriginal people have scored poorly. But, reversing the procedure, European-descended Australian perform quite poorly on tests grounded in Aboriginal culture.

As Rock & Price (2019) argue: “Although standard cognitive processes are similar for all humans, social and environmental differences influence the way cognitive processes are engaged, resulting in different patterns of abilities across cultures. Various studies have highlighted such differences for Aboriginal people. For example, compared to non-Aboriginal Australians, remote Aboriginal people differently approach spatial relationships; and may rely more on visual skills when approaching numerical or memory tasks. Identifying impairment requires understanding which skills are needed for normal function in a specific cultural context” (Rock & Price, 2019, p. 1)

Consider the test shown on the right. How well would you be able to match the hunting tool with the animal hunted?

These points underline Anastasi's argument that intelligence is related to how well people can survive and manage in a particular culture or environment.

Raymond Cattell (1905-1998), John Horn (1928-2006), and the Fluid vs. Crystallized Intelligence Model

The British American psychologist, Raymond Cattell and his student at the University of Illinois, John Horn, developed a psychological model of intelligence that contrasts fluid (Gf) and crystallized (Gc) intelligence to explain general intelligence more precisely. As Plucker & Esping (2018) describe:

"...general intelligence is actually a conglomeration of perhaps 100 abilities working together in various ways in diﬀerent people to bring out diﬀerent intelligences. Gf-Gc theory separates these abilities broadly into, ﬁrst, two diﬀerent sets of abilities that have quite diﬀerent trajectories over the course of development from childhood through adulthood. Fluid abilities (Gf) drive the individual's ability to think and act quickly, solve novel problems, and encode short-term memories. They have been described as the source of intelligence that an individual uses when he or she doesn't already know what to do. Fluid intelligence is grounded in physiological eﬃciency, and is thus relatively independent of education and acculturation. The other factor, encompassing crystallized abilities (Gc), stems from learning and acculturation, and is reﬂected in tests of knowledge, general information, use of language (vocabulary) and a wide variety of acquired skills. Personality factors, motivation and educational and cultural opportunity are central to its development, and it is only indirectly dependent on the physiological inﬂuences that mainly aﬀect ﬂuid abilities"

C-H-C Theory of Intelligence (Cattell-Horn-Carroll)

In 1993 John B. Carroll (1916-2003), a well-known psychologist teaching at the University of North Carolina (Chapel Hill), published a massive re-analysis of more than 400 data sets of cognitive ability scores. It was titled Human Cognitive Abilities: A Survey of Factor Analytic Studies. On the basis of his research, he proposed to extend the Cattell-Horn model of intelligence by arguing that there are actually three levels or strata of intelligence (see the model below). This model has become probably the most widely influential approach to understanding intelligence today by psychologists and those who develop tests of intelligence.

This model argues that "g" (general intelligence) is composed of a set of broad abilities as listed above. Each of these abilities rests, in turn, on a narrow or more distinct set of abilities. Hence C-H-C Theory of Intelligence involves

Stratum I -- the speciﬁc or narrow level; more speciﬁc factors grouped under the Stratum II factors.

Stratum II -- the broad level; eight factors, including ﬂuid intelligence, crystallized intelligence, general memory and learning, broad visual perception, broad auditory perception, broad retrieval ability, broad cognitive speediness, and processing speed

Stratum III -- the general level; general intellectual ability, similar to "g"

A Brief History of Intelligence Testing

A. Binet-Simon Scale of Intelligence (1905)

The French psychologist Alfred Binet (1857-1911) was the director of the Laboratory of Experimental Psychology at the Sorbonne University in Paris. In 1904 the French government turned to a professional group of child psychologists (to which Binet belonged). They were asked to create a way of identifying children who had intellectual deficiencies and needed alternative educational approaches (rather than what students experienced in general elementary schools). Together with his student at the Laboratory, Théodore Simon (1873-1961), the pair created a testing mechanism consisting of 30 tasks of increasing difficulty or complexity which was first published in 1905 and revised twice afterwards.

As Plucker & Esping (2018) describe, “The easiest of [the items on the scale] could be accomplished by all children, even those who [had severe intellectual deficiencies]. Some of the simplest test items assessed whether or not a child could follow a lighted match with his eyes or shake hands with the examiner. Slightly harder tasks required children to point to various named body parts, repeat back a series of 3 digits, repeat simple sentences, and to define words like house, fork or mama. More difficult test items required children to state the difference between pairs of things, reproduce drawings from memory or to construct sentences from three given words such as "Paris, river and fortune." The hardest test items included asking children to repeat back 7 random digits, find three rhymes for the French word obéisance and to answer questions such as "My neighbor has been receiving strange visitors. He has received in turn a doctor, a lawyer, and then a priest. What is taking place?" (Originally found at https://www.intelltheory.com/binet.shtml; the current entry at Wikipedia provides a good overview of this test)

These tasks were evaluated with a set of 50 children with 10 children in each of 5 age groups. These children had been selected by teachers for being "average" for their age. Binet and Simon looked to find at what age most children could successfully complete each task. The ability of average children to get tasks right or correct produced an estimate of the child's mental age (MA). When tested with other children, the scale Binet & Simon developed allowed them to calculate what we know as an intelligence quotient (IQ) based upon the child's MA compared to their actual or chronological age (CA). The calculation of the original IQ was simply MA/CA. Here are examples:

8-year-old child who has a mental age of 10: 10/8 = 1.25

8-year-old child who has a mental age of 6: 6/8 = 0.75

5-year-old child who has a mental age of 3: 3/5 = 0.60

5-year-old child who has a mental age of 9: 9/5 = 1.80

B. Stanford-Binet Intelligence Scale

The Stanford University cognitive psychologist, Lewis Terman (1877-1956), adapted the Binet-Simon Scale for American children and assembled a large group of children to serve as a normative test sample. He expanded the number of items on the scale and sought not only to identify individuals with intellectual disabilities, but also intellectually superior children. The Stanford-Binet tests were originally published in 1916 and, over the years, have been revised four times. While not used as widely in the 21st century as it was in the last century, the 5th edition of the Stanford-Binet is employed by psychologists and neuropsychologists for assessment of both intelligence and cognitive difficulties.

Eventually Terman adopted the suggestion of the German psychologist, William Stern, to calculate the IQ as (MA/CA)x100. In this way, then, a 6-year-old child with a mental age of 6 would have an IQ of 100, but an 8-year-old child who has a mental age of 6 would have an IQ of 75.

Terman used the Stanford-Binet scale to identify very gifted children in California for a longitudinal study of intelligence and its impact upon careers and productivity in life. The results after 35 years seemed to demonstrate that high IQ individuals would achieve great things in life. But, critics noted that the sample consisted of mostly upper middle class White children whose socioeconomic status itself could have predicted much of what Terman's research found.

I will note that Terman himself has increasingly become a figure of great controversy because of his belief in eugenics and, frankly, his outright racism. We do not have time though to explore those aspects of his work and beliefs.

C. Wechsler Adult Intelligence Scale (WAIS)

The Stanford-Binet is a scale that, in the first half of the 20th century, allowed the assessment of the intelligence of children. It was not particularly useful when a testee became either an adolescent or an adult. Indeed, the whole notion of the IQ doesn't make too much sense when, for example, an adult of 25 years of age is tested. If that person's CA is 25, what are the tests or tasks that would measure the testee's MA? What is a MA of 25 or 39 or 62?

David Wechsler (1896-1981), Chief Psychologist at Bellevue Psychiatric Hospital (NYC), was dissatisfied with the Stanford-Binet and its ability to help him and others understand the patients whom he worked with. He criticized the Stanford-Binet as unhelpful among adults. So, in 1939 he created the Wechsler-Bellevue Scale which was the forerunner of the Wechsler Adult Intelligence Scale (WAIS) that first appeared in 1955. Over the years since, it has been revised four times and the latest version, the WAIS-5 was published in late in 2024. Like the Stanford-Binet, the Wechsler Intelligence Scales are individually administered tests consisting of a range of individual subtests. The Wechsler tests take between 1.5 to 2.5 hours to administer.

Wechsler brought several innovations to intelligence testing including the "Deviation IQ" and measuring multiple scales (or forms) of intelligence.

Deviation IQ. Rather than using the notion of mental age, Wechsler argued that intelligence at all ages could better be measured in relationship to the "normal curve". Many, even most human qualities or traits when measured display a pattern that is summarized by the normal curve, that is, most people measure near the average and fewer people at the extremes. For intelligence this means that most people are average but a few are very very intelligent and a few have serious intellectual deficiencies.

So, in constructing the WAIS, Wechsler threw out the notion of CA. Rather, he defined the average IQ as 100 points and applied what is called in statistics a "standard deviation" of 15 points. The resulting distribution of intelligence scores are shown in the figure on the right. A person's IQ is thus measured by how far their performance on the WAIS is different or deviates from the average of 100.

Most people (68%) have IQ scores between 85 and 115 and almost all people (95%) fall between 70 and 130. Only about 2% of people have IQ scores above 130 and about 2% score below 70.

Multiple Scales of Intelligence. The Stanford-Binet offered only a single IQ number as an estimate of an individual's intelligence. In developing his adult scale, Wechsler distinguished between abilities that were based in language and those based in perceptual processes. Thus, in addition to calculating a "Full Scale IQ" (FSIQ), the battery of individual subtests in Wechsler's scale were used to calculate two further scores: Verbal IQ (VIQ) and Performance IQ (PIQ). All three of these individual IQ scores were calculated using an average of 100 and a standard deviation of 15.

Wechsler also researched and published a version of his test battery for use with children. The Wechsler Intelligence Scale for Children (WISC) was first published in 1949 and has been updated four times. The current WISC-V appeared in 2012. The first two editions of the WISC offered the same general scores as the WAIS: FSIQ, VIQ, and PIQ. However, advances in statistical analysis and the availability of more data began to suggest that rather than two, there were at least three or four general domains of intelligence. These findings were also supported by the work of John Carroll mentioned above. As a result, the current WISC-V provides not only a FSIQ, but five additional "Primary Index Scores" involving Verbal Comprehension, Visual Spatial Ability, Fluid Reasoning, Working Memory, and Processing Speed. Similarly, the WAIS-IV provides both a FSIQ and four additional "Primary Index Scores" including Verbal Comprehension, Perceptual Reasoning, Working Memory, and Processing Speed. In 2024, the 5th edition of the WAIS (WAIS-V) was published.