What is I.Q.? (An IQ FAQ)

In his January 16, 2007 Wall Street Journal op-ed, Charles Murray badly distorts the meaning and implications of I.Q. tests and scores. Let's more scrupulously review what I.Q. is, and what it tells us.

What is I.Q.?
I.Q. is a largely effective method of ranking people according to their analytic, or "academic," intelligence. I.Q. scores are highly predictive of how well children will do throughout their school career, and how they will score on other academic tests. The scores also correlate moderately well with future social status and income (Jencks, 1979; APA, 1995).

What do I.Q. tests measure? 
I.Q. tests measure a collection of abilities collectively known as "symbolic logic." The most well known test, Stanford-Binet 5, specifically measures Fluid Reasoning, Knowledge, Quantitative Reasoning, Visual-Spatial Processing, and Working Memory. All I.Q. tests are restricted to pencil-and-paper and do not measure creativity, or what some psychologists call "emotional intelligence" or "practical intelligence."

What is the specific meaning of an I.Q. score?
An I.Q. score reveals how an individual's test performance compares to the rest of his/her age group. Scores are weighted so that the average results always equal 100.

- An I.Q. score of 100 means that 50% of the people in your age group scored better, and 50% scored worse.
- An I.Q. score of 85 means that 84.13% of the people in your age group scored better, and 15.87% scored worse.
- An I.Q. score of 130 means that 2.28% of the people in your age group scored better, and 97.72% scored worse.

What does it mean that I.Q. scores are impressively stable?
Most individuals' I.Q. scores will change little over time (Moffitt et al, 1993), indicating that the test has managed to pinpoint some aspects of a person's "fixed" intelligence -- mind-brain abilities that are inherently a part of that individual and are unlikely to be significantly improved or degraded over time. These basic skills include apprehending, scanning, retrieving and responding to stimuli. The faster one's brain can perform these tasks, the higher one's I.Q. score will be (APA, 1995). Further, a significant portion of these abilities seem to be inherited, as demonstrated by studies of twins separated at birth (Bouchard et al, 1990).

Does I.Q. stability mean that people cannot become smarter and wiser over time?
Far from it. I.Q. stability merely means that, among children who are exposed to roughly equal educational resources, overall rankings of academic intelligence (and academic success) are unlikely to change much over time. If we rank 100 people in 3rd grade according to academic performance, and then wait eight years and rank that same group in 11th grade, the rankings won't be exactly the same but will strongly resemble one another. (Consider: should this be much of a surprise -- that people by and large tend to remain in the same intellectual pecking order as they grow up?)

Fortunately, several powerful trends undercut the significance of the above trend:

1. Almost all human beings, of course, learn tremendously and grow in their skills over time, even if their academic ranking doesn't much change. From a 1995 task force report issued by the American Psychological Association:

"It is important to understand [that] a child whose IQ score remains the same from age 6 to age 18 does not exhibit the same performance throughout that period. On the contrary, steady gains in general knowledge vocabulary, reasoning ability, etc. will be apparent. What does not change is his or her score in comparison to that of other individuals of the same age."

2. Nothing about I.Q. implies any sort of inherent limit, or "ceiling," on the human ability to learn. I.Q simply shows that, as groups grow intellectually, they tend to grow at the same rates. There is plenty of evidence, for example, that schooling raises overall academic intelligence (Ceci, 1990; Darlington, 1995). There is also evidence that most human beings are not reaching their cognitive or academic potential. (Ericsson, 2003; Bartlett and Byrd, 1980). Better schools can raise the level of learning for nearly all students.

3. Any individual can buck a trend, and individuals' I.Q. scores have been known to change as much as 18 points over time (Jones & Bayley, 1941). It is important that we do not confuse "heritability" with "heredity." Scientists have shown that genetics accounts for some portion of the variation of intelligence throughout the human population -- but that is not remotely the same thing as showing that people inherit certain genes that trap them in a certain level of intelligence.

4. Most importantly, perhaps, I.Q. only measures academic intelligence, missing entirely other types of intelligence.

What do I.Q. tests miss?
(How much time do you have?)

Harvard's Howard Gardner: "The tasks featured in the I.Q. test are decidedly microscopic, are often unrelated to one another, and . . . are remote, in many cases, from everyday life. They rely heavily upon language and upon a person's skill in defining words, in knowing fact about the world, in finding connections (and differences) among verbal concepts . . . . An individual can lose his entire frontal lobes, in the process becoming a radically different person, unable to display any initiative or to solve new problems -- and yet may continue to exhibit an I.Q. close to genius level. Moreover, the intelligence test reveals little about an indivdual's potential for further growth." (Frames of Mind, p. 18).

Yale's Robert Sternberg, on the domain of "practical intelligence" not measured by intelligence tests: I.Q. tests "tend to (a) have been formulated by other people, (b) be clearly defined, (c) come with all the information needed to solve them, (d) have only a single right answer, which can be reached by only a single method, (e) be disembodied from ordinary experience, and (f) have little or no intrinsic interest. Practical problems, in contrast, tend to (a) require problem recognition and formulation, (b) be poorly defined, (c) require information seeking, (d) have various acceptable solutions, (e) be embedded in and require prior everyday experience, and (f) require motivation and personal involvement." (1995 APA report).

Also:
- I.Q. scores have a weak correlation with nonacademic intelligence and with performance in everyday tasks in other cultures. (Miller, p. 292).
- I.Q. scores do not identify the most successful and creative artists or scientists (Taylor, 1975, pp.1-36).
- Brazilian street children who have failed math in school can still be savvy in street trades -- they use  methods altogether different from pencil-and-paper math. (Carraher, Carraher, and Schliemann, 1985). (Similar results with California shoppers (Lave, 1988) and harness race wagerers (Ceci and Liker, 1986)
- I.Q. does not distinguish the best chess players from others... (Doll & Mayr, 1987)

***

In his op-ed, Murray equates "I.Q." with "intelligence," which is strongly misleading. He then asserts:
- "Our ability to improve the academic accomplishment of students in the lower half of the distribution of intelligence is severely limited. It is a matter of ceilings."
- "We can hope to raise [the grade of a boy with an I.Q. slightly below 100]. But teaching him more vocabulary words or drilling him on the parts of speech will not open up new vistas for him. It is not within his power to follow an exposition written beyond a limited level of complexity...[he is] not smart enough."
- "Even the best schools under the best conditions cannot repeal the limits on achievement set by limits on intelligence."

Each of these statements is demonstrably false and spreads a dramatically simplified and distorted view of human intelligence. Policy goals aside, it is critical that we present as complete a picture as the current evidence allows.

Comments

[Ray Cheng]

I cannot speak for Charles Murray, but I don't think he would take issue with anything you've written in your above post, apart from your specific assertions that he distorts and is misleading. You write "Each of these statements is demonstrably false and spreads a dramatically simplified and distorted view of human intelligence" without going on to demonstrate their falsehood. Quite the contrary: in your post, you accept IQ as a measure of academic intelligence, and go on to say that IQ is remarkably stable over time. It follows that "Our ability to improve the academic accomplishment of students in the lower half of the distribution of intelligence is severely limited," as Murray notes. Of course, he is simplifying somewhat (he has to with limited space), but this is a far cry from distortion. You go on to write "Policy goals aside, it is critical that we present as complete a picture as the current evidence allows." Splendid idea. Why don't you get in touch with Murray, tell him what you're writing about, and get his input? You might end up agreeing with him -- and if you don't, you'll strengthen your position.

[pawnking]

Re-reading your post, the first thing you state is that 'I.Q. is a largely effective method of ranking people according to their analytic, or "academic," intelligence.'

Mr. Murray's article states that "Half of all children are below average in intelligence." He is clearly stating this in context of academic intelligence. He discusses IQ in relation to academic performance.

If Mr. Murray is being strongly misleading by equating "IQ" with "Intelligence," please explain how your definition is distict from his? I am not certain that I understand your case. It seems that your first point agrees with him; that in the realm of academia, IQ does indeed equal intelligence.

[David Shenk]

No, Ray, it does not follow. Even assuming, for the sake of this debate, that I.Q. is 100% stable -- that is, that your academic ranking among your age group could never change -- it does not follow that we can't raise the academic accomplishments of everyone in our age group, including you. They are two separate things. The first, I.Q., is a question of whether or not you can rise or fall in your academic ranking with others. The second is how much you and your age group can learn from year to year to year. If your best friend in 5th grade has an IQ of 130 and you have an IQ of 80, and you both stay in school through high school, your friend will presumably learn more than you. But that doesn't mean you can't learn a whole lot.

[David Shenk]

pawnking:

First, please realize the logical circularity of "half of all childen are below average in intelligence." It is a meaningless statement, the equivalent of saying: "the more m&m's I eat from the bowl, the fewer m&m's there are in the bowl." Unless everyone has exactly the same intelligence, there will always be an average, and 50% of any group will always be below average. That's inherent to the definition of "average."

Second, we all know there is much, much more to any child's schooling than pencil-and-paper, analytic, "academic" knowledge. The lives and careers of artists, athletes, actors and politicians also develop there -- many of them perfectly mediocre test-takers. Schools are supposed to teach people to think in a curious, skeptical sophisticated way. They can, and ought, to develop all aspects of a person's intelligence. Great schools do, and it is not inappropriate to aspire to have more of these schools.

Thirdly, as I just reiterated in my comment above, it does not follow that just because people retain relatively stable academic rankings within their age group, that the whole boat can't rise with the tide.

Fourthly, Murray pins much of his argument on an academic "ceiling" which does not exist.

[projectshave]

I can see how Murray is seduced by IQ scores. High IQ leads to excellent grades, which leads to top-tier colleges, which leads to high paying/status jobs. Therefore, high IQ --> success. However, I think that once you are on the right academic gravy train, it's hard NOT to land in a cushy middle-class job. Unfortunately, those same academic skills (memorize & regurgitate) don't help in most careers. Non-IQ skills become vastly more important: creativity, social skills, etc. That's why Einstein didn't make zillions from his refrigerator invention. He's brilliant, but in a different way.

[pawnking]

I think that I understand, David. When you state that it is highly misleading to equate IQ with intelligence, you are attacking the arguement as a whole, not the semantic differences between one and the other. If I understand your arguement correctly, you are stating that focusing on IQ solely is a poor way to use your educational resources. Am I correct?

Thanks for the clarificaiton.

[Ray Cheng]

It does follow, if you understand Murray's statement to refer to the performance of the bottom half relative to the top half. Even so, I don't see where you've proved he's trying to distort, mislead, or be unscrupulous.

My more significant point is that Murray's Op-Ed didn't arise from a vacuum. He has research to back up his claims. Since you and he evidently disagree I suggest you contact him and get to his sources.

More suggested reading: Christopher F. Chabris, "IQ Since 'The Bell Curve.'"

[David Shenk]

pawnking: I apologize, but I don't think I quite understand your question. Let me try to restate and tell me if this satisfies you: If Murray was writing about analytical intelligence, and made that clear from the outset, and decided to equate "I.Q." and analytical intelligence, that would be fine. But he's leaving the very strong impression throughout his piece that I.Q simply equals intelligence, which is extremely misleading to the general public. I think if you took his op-ed to 100 people on the street, 99 of them would get the impression that I.Q.=intelligence.

ray: Respectfully, you're misstating Murray's big point. It is not simply to argue that the bottom half will never catch up to the top half. (That wouldn't be much of an op-ed). His point is to say that there are severe limits on what the bottom half can learn, period. The relative stability of I.Q. rankings has absolutely nothing to say on whether individuals in the bottom half face some sort of intellectual ceiling. I'm not aware of any evidence that such a ceiling exists, and I have cited specific evidence to the contrary. If you have any specific evidence for Murray's argument, I'd love to see it.

[Ray Cheng]

David,

* On Murray, okay, I can offer nothing more on this front, and will read at least some of your cited references and try find more that might be relevant.

* With respect to Ericsson et al., I re-read some of the available papers online, and I need to backtrack from my criticism somewhat, but it's not clear to me how much. In one of his papers he makes the point that he is talking about expert performance, rather than performance that is so exceptional (in the sense of rare) that it cannot be reproduced and measured reliably. So he is perhaps talking about (my words, not his) performance at the level of national chess masters, rather than the level of Kasparov and Fischer. But then, his readership doesn't always heed that distinction, and so we get talk about geniuses born vs. made, when the term genius would seem to fall outside the scope of his findings. There is much I must re-evaluate with this in mind.

* Just a fun example of genius. There's a fellow named Noam Elkies, who has enough talent for three lifetimes (I can't get past how unfair that is - some of us don't have enough for one). After completing a degree from Juillard (cello, I believe), he went on to get his Ph.D. in Mathematics at Harvard, where he remains today as a professor. In the mean time, he earned his chess master title, having contested fewer than a hundred games. On a visit to Israel, he found out by chance that the world chess solving competition was being held nearby. Without any preparation, he entered, and won the world championship. (Sorry, I don't remember the year.) He might as well win an Olympic gold medal in the decathlon, too.

[Ray Cheng]

David,
It seems to me that when Murray speaks of academic ceilings, he is not talking about an absolute inability to learn anything beyond a point, but rather the inability to learn certain kinds of things. He mentioned higher levels reading comprehension as something that would demand more ability than some folks have. (I may have just run into my own limitations.) But didn't you speak of an academic ceiling of that type yourself, something about never becoming a mathematician or soccer player?

[David Shenk]

Ray: Sorry if I wasn't clear about myself. I was trying to articulate that I know how powerful the idea of the skill-ceiling is -- even for people who are familiar with the slow process of becoming good at something else. Anyone with mediocre skills at something (like me at soccer) looks at a master in a field and says, "Wow -- I could never get there." What in reality is many, many years and many thousands of work hours away is experienced by the individual as being a *world* away. But it's not. Very, very far is not the same as beyond reach.

What I'm trying to accomplish with the book is to help people see and understand that long, slow process of becoming great so that they can get past that feeling that it is completely beyond their capability.

That is also, in my view, the virtue of the expertise movement. They are making the invisible visible, using scientific technigues to *show* how people get from mediocre to great. It's unfortunate when it gets boiled down to cliches like "10 years/10,000 hours." From where I stand, these expertise researchers are not resting on a few studies as proof that talent doesn't exist. And, as you have just indicated, they don't claim to have the complete handbook on where genius comes from. Rather, they rightly feel that they are contributing crucial pieces of the puzzle. Genetic expression is another important piece of the puzzle. Personality genes (Matt Ridley, Nature via Nurture, pp. 82-83) may be another important part.

People do have differences in their brain, some of which may well be present at birth. By school age, people do fall into an academic pecking order that may not change a whole lot in their lifetime. In population studies, people with 130 I.Q. may rarely do worse academically than people with a 95 I.Q. But none of that suggests a ceiling on what a particular individual is capable of learning, and I would love to see any research that does. Ericsson and others seem to be showing the exact opposite: that individuals with average capabilities have no practical limits -- that it's only a matter of time, training, resources, etc. I'm sure Ericsson would be as interested as I am to see contradictory evidence.

[Ray Cheng]

David, Elsewhere in your blog I have already posted links to critics of these expertise studies, and I find them convincing, especially with regard to the viewpoint "individuals with average capabilities have no practical limits -- that it's only a matter of time, training, resources, etc." Try changing an individual's IQ from 95 to 130, with unlimited time, and your choice of training and resources. There is no known means to do this. Not even remotely. Why would you believe that any person with an IQ of 95 (or 105) can become a mathematician, which arguably demands an even narrower and deeper set of cognitive skills? By mathematician I mean someone who can make a credible living teaching mathematics at the college level and producing publishable research -- not necessarily genius performance. Now, I'm sure that your IQ is much higher than average, but I doubt very seriously whether you could become a mathematician, even if we could turn the clock back to your early childhood. I've seen many, many similarly bright people fail at it. And not for lack of time, training and resources.

[Ray Cheng]

* Here it is, once again, a balanced and generous critique of Ericsson: http://www.smallgraymatters.com/2006/08/15/13/ Please read it.

* Not too long ago I saw a feature on PBS about an autistic savant named Daniel Tammet. It showed him performing incredible feats of computation, memory, and language acquisition. He picked up the Icelandic language in a single week -- well enough to go on Icelandic TV and converse (including jokes!) with his hosts. What makes Daniel different from other autistic savants is that he can describe, to some degree, what goes on in his head. Multiplying huge numbers, for example, is utterly effortless for him - he just sees the numbers transformed into shapes, colors and movements, and then answer pops out. (He has a high degree of synesthesia.) There is a great scene in the documentary in which Daniel (as a test) is asked to make sculptures of randomly chosen numbers out of Play-Doh. Indeed, a week later he reproduces the same shapes with the same numbers. He can recognize prime numbers instantly and without effort: they appear "smooth" in his mental landscape. He can't stand big cities; the tall buildings keep triggering the number "9" in his head, and that proves too distracting. Beaches are also a problem, since he keep getting driven to count grains of sand. Here is a link to his Wikipedia entry: http://en.wikipedia.org/wiki/Daniel_Tammet.

*

[alexander9904]

I am posting in hopes of gaining further understanding of IO testing in specific areas. My son has been tested as NVLD (comparible to asperger's) in which he is on each end of the spectrum of the bell curve, at both 70's and upwards of 138. He is high superior in the area of fluid reasoning. I do not, however understand how to break that down into specific day to day activities, in which I can help develop his strengths. Any information would be of help. Thanks.

[wellwisher]

FAO alexander9904
There's a recent book by Maggie Mamen on Non-verbal learning disabilities and their clinical subtypes:
available at http://www.specialneeds.com/books.asp?id=12666. Byron Rourke has studied NVLD for many years. His website is http://www.nld-bprourke.ca/ In the UK NVLD does not exist (it gets lost in developmental coordination disorder, known here as dyspraxia), which is why I refer you to Mamen.
It's important not to treat your son for problems that he does not have. In 1994 my WAIS results were VIQ 130, PIQ 97 (PSI 4th centile)...they are quite different now e.g. VIQ < PIQ, PSI 18th centile.

[Gracie]

I recently took an IQ exam. I scored 127 on it. Does this mean that I am "average", above average, or below average? I don't quite understand how the whole thing works, to be honest with you. I also found out that I am in the 100th percentile in linguistics and visual areas. I am in the 80th percentile in mathematics and 50th in logistics. Please, explain all this mumbo jumbo to me! Thank you ever so much.