Actually it gets more complicated than cultural bias Frank, in fact there are certain supposedly "culture free" tests, like Ravens progressive matrices that appear to be increasing IQ at a much greater rate than vocabulary or arithmetic, which have hardly changed at all.

This is likely because of what our society reinforces certain skills that can help to increase those parts of the tests. For example, it's thought that video games can help to develop and reinforce certain spatial skills, eg. block manipulation. Society itself is constantly reinforcing certain aspects, often away from the utilitarian and towards the abstract. In fact, most of ancestors of 100years ago would score exactly as much as the supposedly "sub-average" populations are scoring today.

However, it is quite clear that while IQ test scores are drifting in a certain fashion, it really does depend on what kind of society you are from as to how well you do on these tests. And the test scales are recalibrated every few years because of this drift.

Amazing isn't it? The more I learn about IQ scores, the less I believe that it even correlates very well with real intelligence.

I think I'll put an extra $1000 on that bet steven, since I'm not really interested in gaining another nut.

BUGSY, FRANK,

I am NOT certain of this which is why I'd bet $1,000, not everything I own plus everything I could borrow.

I know about the Flynn effect and have read the Scientific American article you linked.

I don't think IQ scores are as loosely linked to intelligence as you imagine. IQ scores are linked to real world outcomes. On AVERAGE, for example, high IQ people out-earn low IQ people even after adjusting for socio-economic status of the family. It is now pretty well established that in the US high IQ African Americans earn as much as their high IQ white or Asian counterparts.

However IQ tests are a blunt instrument as are ALL examinations. We do need something better, As the cost of fMRI technology drops and the resolution increases we seem to be getting that. My guess is that soon, maybe by 2015, we shall be able to measure brain function the way we measure lung or kidney function today.

When it comes to intelligence and brain function generally we are at the stage of pre-X-ray medicine. We have to guess what's going on inside. fMRI and other technologies are giving us the X-ray capability we need.

To get back to my original point about divergent evolution. We simply do not yet know how many genes are involved in intelligence and how they interact with the environment. What we do know is that the influence of genes on intelligence is non-zero. Give two people optimum nurture and one will emerge smarter than the other.

And we also know that some genes that are involved in brain development have been subject to selective pressure quite recently. For an example see:

http://www.sciencemag.org/cgi/content/abstract/309/5741/1720

All these factors lead me to believe we shall probably, not certainly but probably find differences in "brain function" between populations and this difference will translate into a difference in intelligence.

But I would be the first to admit I could be wrong. We'll just have to wait and seen what happens.

Just to clear something up.

My bet on intellectual differences between population groups, "brain function" differences if you like, is one I'd be happy to lose.

I think the evidence is beginning to point in that direction. For me the evolutionary argument is the most compelling. What did keep the evolution of brains synchronised for 3,000 generations?

But I wish it weren't so.

My hope is to be proved wrong.

But, in the end, the world is under no obligation to be as we would wish it to be.

OK, all you IQ squabblers, here's a thought for you.

Some years ago, I experimented with the commonly used IQ tests, gathering together a collection of them from various public and private sources.

I then went through them, one a day, every day for two weeks.

At the end of those two weeks my IQ had apparently "increased" by more than twenty points. In fact, on reviewing the results, it was clear that in no test had I performed worse than the last one - an unbroken line of progressive improvement

It was a bit like doing cryptic crosswords. The tests use a "language" that becomes increasingly easy to work with as you become familiar with it.

For example, when I began doing the Times crossword, I rarely finished it. Now I can generally complete it within thirty minutes. I haven't become any smarter, just more familiar with the process. This is easily proven by giving me a different style of cryptic crossword, I am instantly back at square one.

IQ tests may be a measure of something, but it most certainly isn't intelligence.

Pericles

No question that practice improves IQ scores. This is a well known phenomenon. It's called "test sophistication." It's one of the reasons I call IQ testing a blunt instrument.

In the US there are actually schools that coach people to do IQ and similar tests such as the SAT and GMAT. They do produce results.

However all test coaches admit that people reach a "saturation" point. Some people saturate at higher levels than others.

So the real test is not what is your IQ score but what is the best you can do after practice.

So how much practice do the Africans get?