Okay, this whole IQ thing is starting to bug me. Let's get a few things straight...
IQ, the so called "intelligence quotient" measures how well you do on IQ tests, and not much else. Despite the name, it does not measure a person's native intellectual ability or capacity.'
Some history is in order...
Back around the begining of the 20th century, some folks noticed something. Kids who scored well on certain tests tended to score well on other tests, and tended to get better grades in school. Now, if your intention is to identify ids who, for whatever reason, are at risk of falling behind in school, you can make good use of this fact. You can give diagnostic tests, identify kids who are at risk and keep tabs on them, and possibly identify their problems.
The identification of this correllation with "intelligence" is based on a few things. One is humanity's wish to keep things simple. Another is a misunderstanding of some particulars of statistical analysis. A third is a misunderstanding of one of the most basic idedas of statistics. And, finally, the people who originally made this identification had some nasty ideas on racism and eugenics that they wanted to push forwards.
For an in-depth discussion of all this, I can point you to a book, The Mismeasure of Man by Stephen Jay Gould. I'll summarize a bouple fo the points in brief...
The first is pretty obvious. We like things simple. Viewing "intelligence" as a single thing that can be ranked with one universal number is simple and attractive (especially to those for whom the number is high). Slap a number on a person - this one is smart, that one is dumb, this one is just that minute fraction smarter than another. You can line them all up, nice and neat, placing a value on each person. We like to be able to put simple values on things, even complext things like people...
The second is less obvious (and hard to get across without diagrams). If you plot out test scores as vectors in a particular way, you suddenly see that there seems to be a natural direction they all seem to be pointing. You put a graph axis in that direction, and use it to extract a measure. Call that measure "g" for "general intelligence". And voila, you have the basis for IQ.
Here's the error - it seems that there's a natural direction. But that seeming does not imply that the axis you laid down actually has a real, physical meaning. There are other choices of axes that are just as useful (if not moreso), from which you can extract other information.
Anyone who plays D&D with minis on a battlemap actually understands this. We tend to lay down a quare grid, oriented North-Sough and East-West. But we know that doesn't really mean that Tordek walks directly north more easily than he can walk east-by-southeast. We can rotate the grid under the figures to whatever is most useful. Heck, we can slide a hex grid under Tordek, and suddenly instead of two preferred directions, he has three (north-south and two diagonals)! The choice of grid is actually arbitrary, done for convenience rather than to show a real preference of the universe.
The third issue is a really basic tenet of statistics - correllation does not imply causation. The fact that we can measure a correllation betwen performance on a test and performance in a classroom does not mean we automatically know the cause of that correllation. Some people felt that the underlying cause was actual personal ability, so they used the terms "general intelligence" and "intelligence quotient". But that was their opinion, not an established fact.
What's in a name? Well, in this case, the names stuck. So, we are left with an old, outmoded idea we cannot shake - that the correllation between test scores says something about innate ability, rather than about environment and upbringing.
The last - the racial biases fo the people involved, I will not address, as that's getting into politics...
One final thing - some people have noted that some IQ tests are "only accurate within a given range". That's true. but it avoids the question - Accurate at what? It's not acuracy measureing intelligence. It's accuracy in predicting how well you will do on similar tests!
