"It would be difficult to start measuring “intelligence” without at least some implicit or intuitive theory of what intelligence is, and from the earliest Greek philosophers to the present day, many writers have enunciated their ideas about the nature of intelligence ... . For Plato, it was the love of learning – and the love of truth; St. Augustine, on the other hand, believed that superior intelligence might lead people away from God.  Thomas Hobbes in Leviathan went into more detail, arguing that superior intelligence involved a quick wit and the ability to see similarities between different things, and differences between similar things (ideas that have certainly found their way into some modern intelligence tests)."

Measurement, however, implies something further: No one would be interested in measuring people's intelligence unless they believed that people differ in intelligence. Many early writers did of course believe this. ... And when Francis Galton published Hereditary Genius in 1869, in which he sought to prove that people differed in their natural abilities, his cousin Charles Darwin wrote to him: “You have made a convert of an opponent…for I have always maintained that, excepting fools, men do not differ in intellect, only in zeal and hard work” (Galton, 1908, p. 290).   (p. 3)

Measuring Intelligence

Galton

Francis Galton had no doubt on this score.

I have no patience with the hypothesis occasionally expressed, and often implied, especially in tales written to teach children to be good, that babies are born pretty much alike, and that the sole agencies in creating differences between boy and boy, and man and man, are steady application and moral effort.  It is in the most unqualified manner that I object to pretensions of natural equality.  The experiences of the nursery, the school, the University, and of professional careers, are a chain of proofs to the contrary. (Galton, 1869, p. 12)

The results of public examinations, he claimed, confirmed his belief.  Even among undergraduates of Cambridge University, for example, there was an enormous range in the number of marks awarded in the honour examinations in mathematics, from less than 250 to over 7,500 in one particular two-year period.  As a first (not entirely convincing) step in the development of his argument that this wide range of marks arose from variations in natural ability, he established that these scores (like other physical measurements) were normally distributed, the majority of candidates obtaining scores close to the average, with a regular and predictable decline in the proportion obtaining scores further away from the average.

Allied to an almost compulsive desire to measure anything and everything, it was perhaps inevitable that Galton should wish to provide a direct measure of such differences in natural ability.  But what measures would succeed in doing this?  In 1884, at the International Health Exhibition held in London, he set up an Anthropometric Laboratory, where for a small fee visitors could be measured for their keenness of sight and hearing, color vision, reaction time, manual strength, breathing power, height, weight and so on. He could hardly have supposed that these were all interchangeable measures of intelligence, and some were surely there simply because they could be measured.  But Galton was a follower of the British empiricist philosophers and argued that if all knowledge comes through the senses, then a “larger,” more intelligent mind must be one capable of finer sensory discrimination and thus able to store and act upon more sensory information." (Mackintosh, 2011, pp. 3-4)

James McKeen Cattell

"A more systematic attempt to measure differences in mental abilities was proposed by James McKeen Cattell ... who published a detailed list of 10 “mental tests” (plus another 40 in brief outline); they included measures of two-point tactile threshold, just noticeable difference for weights, judgment of temporal intervals, reaction time, and letter span.  Cattell did not claim that this rather heterogeneous collection of tests would provide a good measure of intelligence – indeed the word “intelligence” does not even appear in his paper.  Once again, it seems clear that the tests were chosen largely because the techniques required were already available.  These were the standard experimental paradigms of the new experimental psychology being developed in Germany, and whatever it was that they were measuring, at least one could hope that they were measuring it accurately.  Although no doubt unfair, it is hard to resist the analogy with the man who has lost his keys when out at night, and confines his search to an area underneath a street lamp, not because he thinks that is where he lost them, but because at least he can see there." (p. 4)

{ clip }

Binet

"It was the Frenchman, Alfred Binet, who solved the problem of devising an apparently satisfactory measure of intelligence.  Although he and his colleague, Victor Henri, had made earlier attempts to measure differences in intelligence, they had not been spectacularly successful ... and it was a commission from the French Ministry of Education that revived their efforts.  The introduction of (nearly) universal primary education had brought into elementary schools a number of children of apparently below average intelligence, who would never had attended school before.  They did not seem to be profiting from normal classroom teaching and were deemed to be in need of special education.  The problem was to devise a quick and inexpensive way of identifying such children.  Binet had little time for the new experimental psychology coming from Wundt's laboratory in Leipzig, and although much less hostile to the associationist tradition of British empiricism, he did not believe that associationism could answer all questions.  Above all, he thought it nonsense to suppose that intelligence could be reduced to simple sensory function or reaction time.  Observation of his own young daughters had convinced him that they were just as good as adults at making fine sensory discriminations, and although their average reaction time might be longer than that of an adult, this was not because they could never respond rapidly but rather because they occasionally responded very slowly – a failure Binet attributed ... to lapses of attention.

For Binet, “intelligence” consisted in a multiplicity of different abilities and depended on a variety of “higher” psychological faculties – attention, memory, imagination, common sense, judgment, abstraction.  Even more important, it involved coping successfully with the world and would thus be best measured by tests that required young children to show they were capable of coping with everyday problems.  Could they follow simple instructions such as pointing to their nose and mouth?  Did they understand the difference between morning and afternoon, and know what a fork is used for?  Could they count the number of items in a display, and name the months of the year (in correct order)?  And so on.  Were these adequate measures of intelligence?  Binet's critical insight was that as young children become more intellectually competent as they grow older, a good measure of intelligence would be one that older children found easier than younger ones; this was particularly relevant for his main task of identifying children who were mildly or perhaps more seriously retarded: The difference between “normal” and retarded children was that the former passed his tests at a younger age than the latter." (p. 5)

{ clip }

Spearman and the Theory of General Intelligence

"Faculty psychology was Charles Spearman's bęte noire.  He abhorred the program that would separate the mind into a loose confederation of independent faculties of learning, memory, attention, and so on.  What was needed was to understand its operations as a whole.  Without knowing about Wissler's experiment, he repeated something very like it with a group of young children in a village school .... He obtained independent ratings of each child's “cleverness in school” (from their teacher) and “sharpness and common sense out of school” (from two older children), and also measured their performance on three sensory tasks.  Unlike Wissler, he did observe modest positive correlations between all his measures: the average correlation between the three ratings of intelligence was .55; that between the three sensory measures was .25, and that between the intelligence and sensory measures was .38.  These were certainly more encouraging than Wissler's results – perhaps because the obvious restriction of range in students at Columbia University lowered Wissler's correlations.  But they were still rather modest.  Undaunted, Spearman argued that this was because his measures were unreliable, and a correction for attenuation had to be applied.  The true correlation between two tests was the observed correlation between them divided by the square root of the product of their reliabilities.  This is of course a standard formula for “disattenuating” correlations between two tests, but in modern test theory, the reliability of a test is measured by the correlation between performance on the test on separate occasions, or performance on one half of the test versus the other.  Spearman had no such information and instead assumed that the reliability of his three measures of intelligence was the observed correlation between them, and similarly for the three sensory measures." (p. 6)

Nickerson, R. S.  (2011).  Developing intelligence through instruction.  In Robert Jeffrey Sternberg & Scott Barry Kaufman (Eds.),  The cambridge handbook of intelligence  (Chapter 6: pp. 107-129).  New York, NY: Cambridge University Press.

Stanovich, Keith, E, West, Richard F., & Toplak, Maggie E.  (2011).  Intelligence and rationality.  In Robert Jeffrey Sternberg & Scott Barry Kaufman (Eds.),  The cambridge handbook of intelligence  (Chapter 39: pp. 784-826).  New York, NY: Cambridge University Press.

Sternberg, Robert Jeffrey & Kaufman, Scott Barry  (Eds.)  (2011).  The cambridge handbook of intelligence.  New York, NY: Cambridge University Press.

Urbina, Susana.  (2011).  "Tests of intelligence".  In Robert Jeffrey Sternberg & Scott Barry Kaufman (Eds.).  The cambridge handbook of intelligence  (Chapter 2, pp. 20-28).  New York, NY: Cambridge University Press. 

This chapter endeavours to respond to the following six (6) questions: 

1. Do intelligence tests have a future?

2. Do intelligence tests really measure human intelligence? 

3. What are intelligence tests? 

4. What do intelligence tests actually do? 

5. What functions or purposes do intelligence tests serve?

6. When and how did intelligence tests come to be?