Landmark research
In 2008 a team of cognitive neuroscientists from Bern, Switzerland and Michigan, United States, demonstrated that a very specific type of cognitive training called the dual n-back can improve intelligence. The article entitled ‘improving fluid intelligence with training on working memory’ can be found here.
This was the first documented experimental effect of brain training on IQ in over 40 years. Why did it take so long to discover an effective training method?
Our short term memory is a big part of the story. We use our short term memory all the time for any storage of information that is short term – perhaps over a matter of 10 or 20 seconds – for example, while remembering directions have have just heard while driving. This is memory by rote and is essentially a passive form of memory storage.
Short term > working memory
But more important than just remembering information by rote is being able to do mental operations on that information – to solve a problem, to figure something out, or reason through something to find an answer. For instance, while figuring out a 15% tip, or how much currency is worth while you are in another country. The ability to hold information in mind for brief periods, and manipulate it mentally is a type of short term memory called working memory. You have to do mental work on the information, not just store it. That is why it is called ‘working memory’.
The capacity of working memory
Most people have a working memory capacity of about 2 or 3 – much less than the classic short term memory span of around 7 (‘magical number 7’) for just storing information without doing any cognitive operations on it such as comparing or ordering or adding.
The working memory-IQ link
People vary widely in their working memory capacity. It is now known that these differences are linked to IQ level. General intelligence – measured by standardized IQ tests – depends on working memory because working memory affects a wide range of complex cognitive tasks besides figuring out a tip, involving reasoning problem solving, and making sense of things. We use working memory every single time we reason, plan and problem solve. Even comprehending long sentences (like the ones in this article) requires working memory!
Working memory and the intelligence behind our IQ level both share the same brain circuitry – part of the frontal cortex of the brain called the dorsolateral prefrontal cortex. (‘Dorsal’ means up, and ‘lateral’ means to the side – hence ‘dorsolateral’). This is one of the most recently evolved parts of the brain.
How to improve IQ?
The logic is simple: If you can improve your working memory capacity by training it directly, you can therefore improve your intelligence level. There is, in technical terms, a ‘transfer effect’ from working memory training to gains in intelligence and IQ.
Intelligence can be improved substantially – as a side effect
The cognitive psychologists at the University of Bern in Switzerland and the University of Michigan in the States, demonstrated that by training on a specifically designed working memory exercise you can increase working memory capacity by over 65% over just 19 days of training.
This improvement in this type of short term memory capacity had a remarkable side effect: a 40% gain in intelligence – as measured by a version of the time limited Raven’s Advanced Progressive Matrices IQ test – one of the most valid and highly regarded IQ tests for culture fair intelligence.
According to this study, the experimenters only allowed the testees 10 minutes on the RAPM “to keep the administration time short”. I therefore suspect that if anything had been improved, it was the speed at which the testees worked on the test, rather than working memory. The RAPM progresses from moderately easy questions to increasingly difficult ones. It is unlikely that in 10 minutes any of the testees were able to complete enough of the test to get onto the more challenging questions.
It perhaps would have been more revealing had the RAPM been administered to the testees in this study in the way that it was originally designed to be administered, i.e. the full amount of time had been allowed for the volunteers to complete the test. That way, it would have been possible to examine which of the more difficult test items the testees answered correctly after practising the dual n-back task, which they had previously missed. Surely this would have been a more appropriate way to test whether their IQ had increased?
Yes, it would have been good to do an unspeeded fluid intelligence test. I am currently doing this study as you can see here: https://markashtonsmith.info/research-experimental/dual-n-back-interventions
However, the working memory benefit – both behaviorally and in terms of known neuroplasticity effects on synaptic connections – as well as established theory about the role of working memory in fluid reasoning, indicates that the effect is not confined to improvements in test taking speed. Also, this software has been selling with a 10-20 point fluid G increase guarantee for nearly 2 years and there have been close to zero cases of failing to get this kind of gain that can’t be attributed just to practice effects.
And of course, there is a high correlation between the speeded matrices tests used in the Jaeggi et al study and the longer RAPM tests – that’s the rationale used in the study itself.
This is very interesting, as is your personal website. I am interested in the effects of this type of training on occupants of the extreme right-hand tail of the distribution curve. I’m interested in all forms of neurotechnology and cognitive enhancement. Sellers of AVE software (such as Transparent Corp) have published information showing the greatest gains for those of below-average ability, and a more modest (5-7 points) increase in IQ for those in the average range.
However, we “outliers” are out of luck, according to them. Supposedly, because we’re already using our mental equipment more efficiently to start with, our mileage may vary. I’m sorry, but I don’t buy it, in fact (as a commenter on my own blog said), I don’t even rent it! One definition of intelligence is “the capacity to acquire capacity” (Woodrow), and by that reasoning the people best placed to acquire even more of what they’ve got are the people who are smart to start with!
Oh, and the sevensigma thing is just a screen handle, by the way. I know there are no tests with enough ceiling to go anywhere near there!