a while back i think i read an article on the straight dope homepage that said that the average person had a 100 person iq score, and that they recalibrate it every few years to keep it that way. anyone have a link to that article? anyone have a link to an article that disputes this claim? thanks
It is not just one article. It is by the very defintion of an IQ score. IQ tests are designed so that a representative sample of the population gets a 100 and the distribution is shaped like a bell curve with a standard deviation of 10. This is important because it means that you can know the precentile a given IQ score falls into just by consulting a z-score chart built on the principles of a normal bell curve.
Even though it is not a true IQ test, the SAT is designed with similar assumptions. As average SAT scores drifted lower as more students took it, the College Board decided to re-center it so that the average score was again 1000. The standard deviation for the SAT is designed to be 100.
I don’t know the article that you are referring to, but this is exactly how IQ was defined.
The idea was to give tests to a lot of people. Let’s suppose there were 60 items on the test. Lets suppose that the average score for the whole population was 38. But when they report the IQ, they move the numbers so that the average is called 100.
By definition, the average IQ is 100.
But IQ tends to creep up, and the average IQ was getting to be 105 as I recall.
IQ tests are renormed to correct for the creep. You might want to google the Flynn Effect which is interesting.
Nitpick – IQs are unlikely to be rising but the numbers obtained from IQ testing are rising.
Nitpick – perhaps you mean to say that intelligence is unlikely to be rising, but IQs, being the numbers assumed to measure intelligence, are rising. 
I disagree highly. Real IQ tests are not something that you can study for so the results indicate something. There are several reasons why IQ score increases may be both real and measurable.
-
Better nutrition. Certainly helps the body, probably helps the brain develop.
-
Better childhood medical care. Major diseases in childhood are relatively rare now compared to all previous generations. A sick child generally suffers a disadvantage even on mental tasks.
-
Smaller families - No one has a very good grasp why but children from smaller families show very real advantages on mental development. The advantage to first-born children is particularly strong. Smaller family sizes mean that the proportion of children who are first-born is increased.
-
Parents have more time and motivation to dedicate to a child’s intellectual development from day 1. Most of the pop-culture theories on how to aid child development are probably crackpot but even if a few work, then that could cause a small rise in IQ scores.
There are several other factors but these are probably the main ones.
I heard a sigma of 15, but I haven’t been able to find a cite. Is that 10 definitely accurate?
10 for Wechsler instruments; I think it’s 15 for Stanford-Binet, but don’t quote me–I’m not at work.
Sorry–hit Enter by accident. Some are SD 10; others are SD 15. I think I reversed it in the last post.
Okay cool, thanks. 
Shagnasty,
Neither of the reasons 1) or 2) that you list are sufficient to explain the Flynn effect. The average I.Q. score has gone up by 30 points in the last 80 years. This would mean that the average person back then was the equivalent of a mildly retarded person today. We know how much nutrition and medicine can affect intelligence, and the improvement over the past 80 years simply isn’t enough to explain the difference. You would have to assume that the average person back then was on the edge of starvation all the time.
I.Q.'s have gone up because this society is a thinking society now. Look at the business world. It has gone from a labor intensive, manufacturing, industrial environment to an intellectual asset environment. People simply use their brains more, which helps them score higher on IQ tests. Also, more kids are in school these days and they take harder classes at earlier ages. This creates better brains.
Also, I imagine the tests have changed since then to better reflect intelligence.
Sorry you are wrong.
The mean for the Stanford Binet IV is 100 standard deviation 16. The new SB V - mean 100, SD 15, which brings it in line with the Wechsler tests which have a mean of 100 and a SD of 15 as well.
The Wechsler subtests have a mean of 10 and an SD of 3.
This is wrong as well, as new tests are released periodically and are renormed (or recalibrated as the OP called it) on the current population. Therefore you can certainly score highly on an old IQ test today, but you will still be scored on a distribution of M=100; SD=15 on a test with current norming. Results obtained on outdated tests are not considered to be valid.
While it is true that one cannot “study for” IQ tests (in the way that affluent kids currently “prep” for the SAT, several studies by Stephen Ceci and Wendy Williams of Cornell University have indicated a consistent correlation between historical, current, and recent school attendance and IQ–including such simple changes as kids scoring higher on IQ tests at the end of the school year than they do following summer vacation.
To me, this challenges any number of assumptions regarding the neutral or education-independent nature of IQ tests, particularly when compared across cultural boundaries.
auliya writes:
> This is wrong as well, as new tests are released periodically
> and are renormed (or recalibrated as the OP called it) on the
> current population.
Oh, come on, you know perfectly well what I meant. Of course I know about renorming. Please don’t assume that I’m ignorant. According to the Flynn effect, if a person today would get a score of 100 on an I.Q. test, he would have gotten a score of (about) 103 on a I.Q. test given 10 years ago. However the tests are regularly renormed so that the average score stays at 100. Since I.Q. testing has been going on for 80 years or so, and the Flynn effect has been going on pretty consistently over that period, a person who would have gotten about 76 on an I.Q. test today would have gotten about 100 on a test given 80 years ago. This means that a person who is mildly retarded today would have scored average on one of the I.Q. tests from 80 years ago.
Here is Flynn’s most recent thoughts about IQ:
Whatever this (real,imaginary,apparent. . .) improvement has been, it’s due to Cecil and the Straight Dope’s war on ignorance.