What is intelligence? - Chapter 13

How is intelligence classified?

Two groups can be distinguished with regard to the classification of intelligence. On the one hand, intelligence is considered as one factor that predicts inter-individual differences and performance. On the other hand, intelligence is divided into different cognitive domains.

The g factor

Galton (1822) suggested that a smart person had a good neural network, while less smart people did have poorly functioning neural networks:  one characteristic of the nervous system was responsible for 'intelligence'. Spearman (1863), one of Galton's followers, concluded that there was a general factor the g factor. He believed that there was also a specific factor (s factor). Within this approach, there is one general intelligence that causes interindividual performance on cognitive tests. Spearman thought that the g factor included mental energy and that this factor was hereditary. Even today there are many people that support Galton. They based their own well-founded empirical theory on Galton's view, such as Jensen (2011) and Deary (2012).

Which theories divide intelligence into multiple factors?

Thurstone (1887) developed factor analysis and concluded that there are seven 'primary mental abilities': verbal comprehension, word fluency, number facility, spatial visualisation, associative memory, perceptual speed, and inductive reasoning. Gardner (1983) was strongly influenced by Thurstone. However, he did not believe in the g-factor and assumed that there are seven mental sources for information processing: each is used for a certain activity. His seven mental sources are a verbal-linguistic, logical-mathematical, spatial-visual, bodily-kinaesthetic, musical, interpersonal, and intrapersonal intelligence. Sternberg (contemporary) recognises the g-factor, but just like Spearman he assumes that there are also specific forms of intelligence: analytical, creative, and practical intelligence.

The core of the intelligence debate

The above-mentioned persons form the core of the intelligence debate: Galton and Spearman versus Thurstone, Gardner and Sternberg: is there one general g-factor or is it a collection of mental properties?

How is intelligence being measured?

The first intelligence test

Binet (1905) (together with Simon) was asked to develop a method that could be used to determine whether a child was or was not eligible for education. Binet developed several tests that were based on what children were expected to learn or to be able to achieve at school at different ages. Binet's test battery was translated into English and formed the 'Stanford-Binet Intelligence Scales'. Terman adjusted some test components in the intelligence quotient (IQ) by multiplying the ratio between chronological age and mental age by 100. Meanwhile, a lot of tests have been developed for both children and adults. Three of them will be discussed.

The Wechsler scales

Wechsler (1896) developed the 'Wechsler Adult Intelligence Scale' (WAIS). The latest version (2012) has fifteen subtests that are divided into verbal and performance subtests. The verbal subtests determine the verbal IQ, and the performance subtests determine the performance IQ. All raw subtest scores are converted to standard scores. There are four total index scores: verbal comprehension, processing speed, working memory, and perceptual reasoning. All four factors are strongly correlated, which highlights the fact that it is primarily the g factor that determines performance.

The difference between crystallised and fluid intelligence

Cattell (1943) introduced a crystallised form of intelligence (academic knowledge, such as vocabulary, general knowledge, and numeracy), and a fluid form of intelligence (abilities that can be used in novel problem situations). Crystallised intelligence hardly deteriorates as we age, whereas fluid intelligence is greatly affected by age. 

The Raven Progressive Matrices

Raven (1938) introduced the 'Raven progressive matrices' (RPM) as a test independent of cultural factors. According to Spearman it measures the g factor in the most unbiased way. The RPM is not a test battery. In the test Raven used a pattern in which a missing piece has to be chosen from among several alternatives. 

The National Adult Reading Test

The National Adult Reading Test (NART) is designed to estimate premorbid intelligence levels. It is a short test that measures vocabulary by asking the test subject to read out a list of 50 phonetically irregular words. The premorbid intelligence level can be estimated based on the test performance. The correlation with the verbal Wechsler IQ is strong (0.8).

The Flynn effect

The Flynn effect states that the average intelligence level increases by around five IQ points per year. This effect is universal, but not equally strong: the Flynn effect can vary from 5 to 25 points (Flynn, 1987).

How does intelligence work biologically?

The biological basis is the most determining factor of intelligence. Nowadays there is more and more supporting evidence for the notion that intelligence (the g factor) is mainly dependent on the functional characteristics of the connections between brain areas (and especially the connection between the parietal and frontal areas).

Is intelligence hereditary?

Jensen triggered the discussion about the heredity of intelligence in 1969 when he suggested that a genetic cause could explain the differences in the performance of whites and African-American children. According to him, there are two types of skills: learning / memory, and abstract reasoning. White and Asian children would perform better in the latter category. A great deal of research has been conducted into the relationship between intelligence and the speed of information processing (with reaction time RT as a measure in particular), which ties in with the existence of a g factor. According to Bouchard (2004), there is a very high heredity for the g factor (0.85). Also a large-scale study in Amsterdam that has been testing twins since 1977 states that intelligence is 80 to 90% genetic. The volumes of white and gray matter that are correlated to intelligence (0.25) are also genetically determined (Posthuma, 2003).

Where is intelligence located in the brain?

In higher-order animal species, the brains are larger in relation to the body, the weight is higher and the number of gyri and sulci is significantly higher. According to Toga and Thompson (2005) there is evidence of certain brain structures that are responsible for the heredity of the g factor. They emphasize the amount of gray matter in the frontal areas. According to Chiang and colleagues (2009), white-matter integrity is also related to IQ scores and is also highly hereditary. According to the neural efficiency hypothesis, more intelligent people have less activation for a particular task because they work more efficiently and therefore do not need more cells (Neubauer & Fink, 2009). Diffusion tensor imaging (DTI) measurement can be used to analyse how efficiently a signal is transmitted. Research by Tamnes and colleagues (2010) shows that DTI data of the left hemisphere is correlated with the verbal IQ and the performance IQ. The level of intelligence thus appears to be strongly correlated with the quality of the fiber pathways. 

What effect does brain injury have on intelligence?

In a study by Duval and colleagues (2008) participants (ranging from 0 to 84 years) were tested for intelligence after a single unilateral lesion. The Kennard principle did not seem to apply here: lesions at a younger age were associated with a lower IQ (while you would expect a higher IQ based on the high plasticity). According to Woolgar and colleagues (2010), lesions within the parietal-frontal network are related to low IQ scores on IQ tests that measure fluid intelligence. A study by Gläscher and colleagues (2010) suggest that intelligence (the g factor) is mainly determined by the connections that integrate working memory, visual-spatial, language, and executive processes.

To what extent is intelligence important in neuropsychological practice?

Why is the current level of importance?

The purpose of an intelligence test is usually to determine one's current abilities compared to peers. In neuropsychological assessment, determination of the patient's current capacities can be important because it provides a background against which performance on other neuropsychological tests can be compared. 

How are abnormalities being established?

An intelligence test is also used for the detection of functional disorders. However, doing this is difficult because the premorbid level is often unknown. This is done by an estimate based on age and gender. Another solution for this are hold tests: tests in which scores do not decrease with age and are also not susceptible to brain damage. Moreover, it is also conceivable that crystallized intelligence will be affected less quickly than fluid intelligence.

Are IQ tests valid in patients with brain damage?

Many IQ tests measure fluid intelligence, which is very sensitive to the effects of cognitive deterioration. The same problem is found with regard to the ideomotor or mental slowness that often occurs with different brain impairments. Using hold tests can solve this problem, but these too are not always valid.