Image Map

Intelligence and its Measures – Part I
by: Brendan Alexander

Author’s note: The topic of intelligence is extremely complex and riddled with misconceptions and disinformation. When writing this, I wanted to give the basic information so everyone is on the same page when I get to the more interesting stuff in part II… So bear with me if this is a little dry.

What the shit is intelligence?

The concept of intelligence is certainly not foreign – it is easy enough to walk into a grade school and ask who the “smartest” student is – but it is a notoriously difficult term to define. Words like clever, quick-witted, street-smart, sagacious, and bright are commonplace but largely uninformative; none of them really yields a clear-cut definition of what it means to be smart. Psychologists have tried to wrap their fingers around the abstract construct of intelligence with varying degrees of success, but voluminous papers and jargon-filled definitions do not necessarily provide a greater understanding. In fact, Sternberg found that psychologists and the laymen both have a similar understanding of intelligence (Sternberg et al., 1981).

One major problem is that the word “intelligence” connotes so much more than pure mental ability. For better or for worse, people often judge themselves and others based on their intelligence, equating it with personal worth (as is illustrated in the oft-heard lament, “I hate stupid people!”) To deal with all the excess baggage, Arthur Jensen, well known as a vocal proponent of general intelligence, has suggested we stop using the word “intelligence” altogether and instead speak of mental abilities (Jensen, 1998). I agree with his idea, and will pick up and explain that convention in part II when things get slightly more technical. In this article however, I will stick with “intelligence” because, as long as we remember to drop the connotative meanings, it makes things a little simpler.

What do the experts say?

As stated above, even the experts do not agree on a single definition of intelligence. Still, it is informative to read some individual definitions to get a feel for their similarities and differences. Robert Sternberg, one of the major researchers in human ability, defines intelligence as “the cognitive ability to learn from experience, to reason well, to remember important information, and to cope with the demands of daily living” (Sternberg, 1988). David Weschler, the designer of the WAIS tests [see below for a WAIS overview], has said intelligence is “the global capacity to act purposefully, to think rationally and to deal effectively with the environment.” In The Journal of Education Psychology, Louis Terman, preferring brevity, says intelligence is simply “the power to think abstractly” (Jensen, 1998).

Generally, most people agree on at least a few things about intelligence; namely that:

    • Intelligence is concerned with the mind and mental abilities.
    • Intelligence is not synonymous with achievement though the two are often correlated.
    • Intelligence is related to problem solving and reasoning ability.
    • The ability to understand abstract relationships plays a strong role in intelligence.
    • Intelligent behavior is adaptive.

Pissing contests

At first glance, the amorphous definition of intelligence seems like a major problem for the study of human ability as a whole, and psychometrics especially. How can we measure something if we can’t explicitly define it? Consider a simple analogy: basketball ability.

No doubt a symposium of coaches, athletes and beer-swigging fans would yield numerous, slightly different definitions of what constitutes basketball ability. Is it shooting skill? Maybe a combination of speed and jumping ability? Quick reaction time perhaps? Can the definition be illuminated by just looking at statistics? Even without an exact consensus, most people can differentiate a good basketball player from a bad basketball player.

Discriminating between basketball ability is even easier if we are able to make the athletes compete in various tests of skills, even if none of the events really gets to the crux of what it means to have great basketball ability. I would wager that testing and rank-ordering 20 people on their height, ability to shoot free throws, and speed would give a fairly good idea of who the better basketball players are.

Intraspecies intelligence is a comparative measure, and to quantify it (or vilify it, to some of the PC crowd) there needs to be some sort of metric — an unbiased measuring stick to figure out individual differences. Enter IQ tests.

History is boring

I dislike history as much as the next guy (unless the next guy likes it) and I don’t want to bring too much history into the article, but a short history of IQ tests can really shed some light on intelligence. Here is the Cliffs Notes version:

Francis Galton has the idea that intraspecies differences extend to mental abilities (probably influenced by the work of his cousin Charles Darwin), and writes a book about it called Hereditary Genius in 1869. He puts forth the idea that mental abilities are inherited, that is, he would take the side of nature in the “nature vs. nurture” debate, a phrase he popularized (Galton, 1874). Galton reasoned the best way to measure intelligence was by quantifying sensory acuity (1883). Another psychologist, Cattell, based 10 different mental ability tests on Galton’s ideas, but it turned out they did not really end up measuring much (Wissler, 1901).

Alfred Binet stepped up and realized that there really needs to be an independent criterion on which to measure intelligence. Binet also noted that as a child matures (and he was only concerned with children at this point) he or she tends to be able to grasp and understand more, and thus is more intelligent. This allowed him to assign a mental age to a test subject based on what he could or could not do.

Binet’s test was widely accepted and used as a measuring stick for all others. His idea of mental age, however, was soon tossed aside because it certainly can’t be applicable to adults. Wechsler just assumed that all adults had a mental age of 16 and came up with the IQs based on standard deviations that are used today [see note 1].

Note 1:Most IQ tests today have an average score of 100 and a standard deviation of 15.

Test Validity

The skeptical reader will probably notice that there seems to be a giant leap of faith at work here, going from intelligence as an abstract property that escapes rigorous definition to something that is being measured by readily administered tests. In the basketball analogy used above it would be fairly easy to see if some particular test really measures what we call basketball ability – examine the performance of NBA superstars (or maybe Argentinean players, ha ha) compared to the performance of mediocre NBA players, compared to the performance of people who have never played basketball [see note 2].

The validity of IQ tests can be looked at with that same idea in mind. Studies have shown that IQ correlates well with occupational success, grades, and salary—all things historically seen as indicative of intelligence (Binet, 1911; Terman, 1916; Jensen, 1998). N.J. Mackintosh mentions an interesting study by Harrell and Harrell in his book IQ and Human Intelligence that shows the average IQ and range of IQ for various professions. This is a condensed version of the complete table:

Lawyer 128 94-157
Teacher 123 76-155
Machinist 110 38-153
Carpenter 102 42-147
Farmer 93 24-147

The most salient data isn’t that lawyers have higher average IQs than farmers, but that the range of IQs generally increases as the average IQ gets lower – an extremely intelligent person can farm, but someone not so intelligent probably can’t be a lawyer. Again, this matches up with the generally accepted definition of intelligence.

Another generally accepted idea about intelligence is that it is invariant over time. In other words, a measure of intelligence should have high reliability. That would imply that the test is measuring some particular attribute instead of just chance. Without getting too technical, IQ tests as a whole are very reliable, so we know they are measuring something.

A quick examination of the WAIS-III

Before getting into some of the more technical and interesting bits about intelligence (and I promise part II will be more fun) it really helps to have an idea about what sort of things are on IQ tests. In addition, if you do not know much about IQ tests it’s probably interesting to find out what sort of items they’re composed of. I have picked the Wechsler Adult Intelligence Scale, 3rd Edition for the examination because it is sort of considered the gold standard now, and its myriad disparate subtests will be useful when I talk about general intelligence in part II of this article series.

The WAIS is an individually administered test that contains fourteen subtests that are divided into a performance and a verbal category (actually, they’re divided into four categories, then grouped together to form the performance and verbal sections, but that is not important.)

The verbal section includes the following subtests:

Information is a general knowledge quiz that covers a wide range of topics from science to politics.

Vocabulary requires the subject to provide definitions to certain words ranging in difficulty from commonplace to obscure.

Comprehension asks about appropriate (i.e. intelligent) actions in certain circumstances, like why it is inadvisable to drink and drive. It also asks about the meanings of certain proverbs.

Arithmetic is a subtest that requires the subject to do various sorts of mental arithmetic problems. Some are word problems, some are simply addition.

Digit Span presents the subject with a series of digits that must be repeated forwards and backwards.

Letter Number Sequencing requires the subject to listen to a series of numbers and letters, then repeat them back, numbers first, then letters, with the numbers in numerical order and the letters arranged alphabetically. For example, the sequence 4 U 9 B A would become 4 9 A B U.

Similarities asks about how two things are similar. For example, an orange and an apple are both fruit.

The performance section includes:

Block Design where, given nine dichromatic blocks, the subject must make a specific arrangement as fast as possible.

Picture Completion, which is a test where one must identify the missing part from a common object, such as a staircase missing a stair.Picture Arrangement, where the subject has to arrange a series of pictures so they form a story.

Object Assembly, which is just a few, simple jigsaw puzzles. This subtest is no longer used to calculate IQ because of its poor reliability.

Digit Symbol, a timed test where the numbers one through nine correspond to an abstract symbol. After briefly studying which numbers represent which symbols the subject is presented with a sheet that has 90 digits and space between each. He or she must fill in the appropriate symbols as quickly as possible.

Matrix Reasoning displays a grid matrix that contains a certain number of symbols and one blank space. The symbols in the matrix follow a pattern. For example, each space in the matrix shows a picture of a shape; looking from the left to the right in the matrix, the number of sides the shape has increases by one. The subject must figure out which symbol should go in the blank space.

Symbol Search contains two different groups of symbols, a target group and a search group. As quickly as possible, the subject must determine if a symbol in the target group appears in the search group.

Closing

In review, we have discussed a little about how intelligence is (or is not) defined, how IQ tests try to measure intelligence, a little about testing history, and we’ve discovered that intelligence can be pretty boring. The next article will look at the positive correlation between mental tests and general intelligence, a little bit about what this means in the real world, how IQ maps achievement, and what it means to possess high general intelligence. I’ll also explain how to convert IQ to penis size. Really. It’s fun.

Note 2. Obviously the idea of what makes a great or mediocre player isn’t defined here. It’s irrelevant really for the discussion but something like points per game, or other players evaluation could be used.

Works Cited

Binet, A. (1911). Nouvelles recherches sur la mesure du niveau intellectual chez les enfants d’ecole. L’Annee Psycologique, 17, 145-201.

Cattell, J.M. (1890). Mental Tests and Measurements. Mind, 15, 373-381.

Galton, F. (1869). Hereditary genius: An inquiry into its laws and consequences. London: Macmillan.

Galton, F.(1874) On Men of science, their nature and nurture. Proceedings of the Meetings of members of the Royal Institution, 227-236.

Galton, F. (1883) Inquiries into human faculty, and its development. London: Macmillan.

“Intelligence and its measurement: A symposium.” Journal of Educational Psychology, 1921.

Jensen, A.R., (1998). The g Factor: The Science of Mental Ability. Westport: Praeger Publishers.

Mackintoch, N.J. (1998). IQ and Human Intelligence. New York: Oxford University Press.

Sternberg, R.J., Conway, B.E., Ketron, J.L., and Bernstein, M. (1981), People’s conceptions of intelligence. Journal of Personality and Social Psychology.

Sternberg, R. (1988). The triarchic mind: A new theory of human intelligence. New York: Viking.

Terman, L.M. (1916). The measurement of Intelligence. Boston: Houghton Mills.

Wissler, C. (1901). The correlation of mental and physical tests. Psychological Review Monograph Supplement, 3, mo.6.

×