In my last post I argued that  brain size among whites in Western countries has been increasing by about 0.44 standard deviations (SD) per half-century, or 0.88 SD over the last 100 years or so, probably because of improvements in the biological environment (i.e. nutrition, disease reduction as scholar Richard Lynn proposed).

Scholar Arthur Jensen suggested one could conservatively estimate gains in “real intelligence” by multiplying gains in brain size by the correlation between brain size and IQ.  Assuming brain size gains of 0.88 SD over the century, and an IQ vs brain size correlation of 0.35, this predicts “real intelligence” has improved by 0.88 SD (0.35) = 0.31 SD over the last 100 years (roughly 5 IQ points).

And yet the Flynn effect tells us tells us IQ test performance has increased by the equivalent of 30 IQ points over the 20th century.  I used to think most of the Flynn effect reflected a real biological gain in intelligence (though not necessarily g) but I now believe it’s 83% culturally acquired knowledge and skills.

In order to measure the part of the Flynn effect that’s real, you need culture fair IQ tests.  Of course there’s no such thing as a perfectly culture fair IQ test but the following tests come close:

1. Block Design
2. Digit Span
3.  Simple reaction time

Block Design is culture reduced because it requires no language, no numbers, no use of a pencil, no columns and rows or deep concentration like on the Raven Progressive Matrices.  It’s basically pure non-verbal reasoning and it’s the kind of fun hands on task that engages the interest and motivation of people from every culture.  Even cavemen living 80,000 years ago could take it.  Some people think kids who play with Rubik’s cubes would have an unfair advantage, but I don’t buy that because practice at one cognitive domain does not transfer to another, not even a very similar one.  On my other blog I wrote:

A Promethean once told a story about how 90% of chess grand masters could not solve a chess puzzle that many high IQ people who suck at chess can easily solve. If I recall, the problem involved putting many queens on a five by five chess board so that all of the queens would be safe. Grandmasters often have ten or more years of education, knowledge and understanding related to chess, yet were not able to solve a chess problem that high IQ chess beginners with chess ratings about 1000 points lower could solve quite quickly. This demonstrates that education, knowledge, and understanding improve specific skills but do virtually nothing to improve one’s ability to adapt when the situation changes and adaptability is the essence of intelligence. Chess grandmasters are experts at playing chess on an eight by eight board with one queen per player, but all that expertise did not increase their intelligence. It didn’t even increase their chess intelligence because they were no longer chess masters when the rules of chess changed. All it improved was their ability to play the very specific form of conventional chess they had practiced.

Similarly, Digit Span also comes close to being culture fair.  Respected blogger Dr. James Thompson writes:

I hope you will agree that this is a simple test, easy to understand, and largely bereft of any intellectual content. All you need is: to know the names of single digits, and to understand the simple instructions and examples given so that you repeat the digits forwards, and in the later version of the test, backwards. In particular, if you can do digits forwards you reveal you know your digits and have some memory, and if you can do a short string backwards you reveal that you have some memory and you understand the idea of repeating digits backwards.

The test is not only bereft of intellectual content, but is also low on cultural content. Once you have learnt digit names you are ready to do the test. I assume that forwards and backwards are concepts understood by all cultures worthy of the name.

Indeed if you replaced digits with nonsense sounds, you might even be able to give this test to a child raised by wolves.  And while people who live in a culture where there’s a lot of memorizing may seem to have an advantage, unless they’ve specifically practice remembering numbers, they’re unlikely to have an edge.  As I wrote on my other blog:

Scholar Arthur Jensen cited a study where people practiced over and over to improve their short-term memory. At first people could repeat only seven digits from memory, but with practice they could improve up to 100 digits. But when asked to repeat letters instead of digits, they were back down to only seven. So all that endless training did nothing to improve their short-term memory span, let alone their overall intelligence. All it improved was the extremely narrow specific ability they practiced.

Lastly, simple reaction time is so devoid of culture that even people with zero education can perform the task in under one second.  And evidence suggests it can’t easily be trained.

So assuming these three tests represent something close to a culture fair battery, it is interesting to ask how much they’ve increased over the last 100 years or so.

According to scholar James Flynn’s book Are we getting smarter?, Block Design on the WAIS has been increasing by the equivalent of 2 IQ points a decade, so on a scale where white Americans today have a mean IQ of 100, whites a century ago would have a Block Design IQ of 80.

Meanwhile new evidence suggests Digit Span scores have been flat, so whites a century ago would have a Digit Span IQ of roughly 100, just like today.

Lastly, a paper by scholar Michael Woodley et al found that simple reaction time was actually faster 100 years ago in Western countries.  Although the paper used statistical inferences to argue that general intelligence was 12 IQ points higher 100 years ago, the raw data showed simple reaction time itself was the equivalent of 6 IQ points higher.

So summing up, 100 years ago, on a relatively culture fair battery, whites would likely score:

Block Design IQ: 80

Digit Span IQ: 100

Simple Reaction Time IQ: 106

Full-scale IQ: 93

I very crudely estimated the full-scale IQ by treating each of the three tests as subtests on the WAIS-IV (two of them are) and then prorating since the WAIS-IV has ten core subtests, not three.

It’s interesting to note that an IQ of 93 (with reference to today’s white mean of 100) suggests real IQ has only increased 7 points over the century, which is very similar to the 5 point increase I predicted at the start of the post from brain size gains.

It’s also interesting that the most g loaded “culture fair” test (Block Design) increased the most, the least g loaded (simple reaction time) decreased, suggesting a good correlation between the “culture fair” Flynn effect and g, which might be expected if it’s caused by brain size.  Though with only three data points, the correlation with g might be meaningless.