Monday, June 09, 2008

Traing working memory increases fluid intelligence (Gf): New research

Training working memory can increase fluid intelligence (Gf). Wow. Hmmmmm?

I've had a number of people forward the following abstract to me. After reading the article I now see why. The article in the Proceedings of the National Academy of Sciences (PNAS) reports that a working memory training intervention produced positive transfer effects in fluid intelligence (Gf). Cognitive ability training research suffers from a paucity of studies that demonstrate positive transfer to other tasks/domains that differ from the training medium. This study also adds additional strong evidence to the link between working memory and Gf.

Cool stuff. A must read. Much has been written about the link between working memory and Gf. Here are some prior related posts touching on the topics of working memory and Gf.

I need to take time to read this study in depth before commenting more. I've only skimmed the abstract at this point in time. Maybe others can read and comment.

Jaeggi, S., Buschkuehl, M., Jonides, J. & Perrig, W. (2008). Improving fluid intelligence with training on working memory. Proceedings of the National Academic of Sciences, 105 (19), 6829-6833. (click to read)


Abstract
  • Fluid intelligence (Gf) refers to the ability to reason and to solve new problems independently of previously acquired knowledge. Gf is critical for a wide variety of cognitive tasks, and it is considered one of the most important factors in learning. Moreover, Gf is closely related to professional and educational success, especially in complex and demanding environments. Although performance on tests of Gf can be improved through direct practice on the tests themselves, there is no evidence that training on any other regimen yields increased Gf in adults. Furthermore, there is a long history of research into cognitive training showing that, although performance on trained tasks can increase dramatically, transfer of this learning to other tasks remains poor. Here, we present evidence for transfer from training on a demanding working memory task to measures of Gf. This transfer results even though the trained task is entirely different from the intelligence test itself. Furthermore, we demonstrate that the extent of gain in intelligence critically depends on the amount of training: the more training, the more improvement in Gf. That is, the training effect is dosage-dependent. Thus, in contrast to many previous studies, we conclude that it is possible to improve Gf without practicing the testing tasks themselves, opening a wide range of applications.

7 comments:

Anonymous said...

you miss the major failure of the study: we do not know for how long the result will last. The history of IQ enhancement is littered with miracle improvements that have vanished a couple of years down the road. In other words, snake oil...and plenty of naive buyers...Sigh.

Ashton said...

I wouldn't call that a failure since it wasn't the study's purpose—besides, they are working on that aspect as we speak.

Torkel Klingberg found a similar phenomenon in one of his studies: http://www.klingberglab.se/pub/Westerberg2007b.pdf

He also found that gains were still remarkably positive after eight months from the end of training. My hunch is that the plasticity related to working-memory training is similar to the plasticity related to musical training. There a severe cortical changes that are now "hardened" in the musician's brain after many years of training. I bet that it will be the same with many years of working-memory training.

Anonymous said...

Klingberg's research is fundamentally flawed with the following difficulty:
1. The g factor is the differential capacity to profit from training.

2. The WM training was a type of training, and therefore, its efficacy depends on the g factor of the subjects SELECTED (N = 3) for training.

My prediction is that if he trains two groups - one of mildly retarded and one of mildly gifted, the latter group would increase its g advantage over the former.

The presumed fact that g could be boosted is quite a different matter from society's main problem: dealing with differences in g between individuals.

What is needed is finding a way to erase IQ differences, not merely to boost IQ undiscriminately (and hence to magnify pre-existing genetic differences).

I bet that nobody will find any time soon a solution to this challenge.

mind evolve intelligence training said...

Interesting thoughts!

I've been watching recent reports of genetic changes in the brain in response to challenging stimuli (aka training) with growing excitement. After reading Jaeggi's research results I went out and built software to allow people to achieve the same results at home. I'm hoping that we'll get some anecdotal data. But if degus can learn to use rakes, I'm confident that the assumption that intelligence can't be trained will one day be consigned to myth...

Want to prove me wrong? Try this for yourself -- it's miraculous.
www.iqtesttraining.com

It follows Jaeggi's method.

Erik said...

It would seem pretty clear that working memory is tied to gF. In the study, the subject's experienced an increase in both. However, the authors were careful to investigate this, and they found that even though there was an increase in working memory, it wasn't enough to account for the entire increase in gF. They hypothesize that the rest comes from an increase in the ability to make fast connections between items.

Anonymous:
If this task just raised everyone's intelligence by a constant amount, sure we'd still have inequalities, but I would argue it's still a better world. Better to have everyone smarter than to stay where we are.

If anyone is interested, my wife and I have written an open source implementation of the dual-n-back training task. It's at http://www.soakyourhead.com/N-Back.aspx . And the source code is at: http://www.soakyourhead.com/dual-n-back-open-source.aspx .

Alan said...

"My prediction is that if he trains two groups - one of mildly retarded and one of mildly gifted, the latter group would increase its g advantage over the former."

FINE. What's wrong with that?

"The presumed fact that g could be boosted is quite a different matter from society's main problem: dealing with differences in g between individuals."

Why is that "society's main
problem"? Who says?

"What is needed is finding a way
o erase IQ differences, not merely to boost IQ undiscriminately (and hence to magnify pre-existing genetic differences)."

WHY?

I think we need was to boost
I.Q. indiscriminately.

PS: there's plenty of unexploited
technologies for steepening the
sides of the bell curve, i.e.
for reducing I.Q. differences,
at least on the low end (i.e. get
rid of most of the very low
numbers). ERASURE of differences
should not be the goal, and
will not be possible in any
case. But getting rid of a lot
of the low numbers is quite
possible.

Alan said...

"My prediction is that if he trains two groups - one of mildly retarded and one of mildly gifted, the latter group would increase its g advantage over the former."

FINE. What's wrong with that?

"The presumed fact that g could be boosted is quite a different matter from society's main problem: dealing with differences in g between individuals."

Why is that "society's main
problem"? Who says?

"What is needed is finding a way
o erase IQ differences, not merely to boost IQ undiscriminately (and hence to magnify pre-existing genetic differences)."

WHY?

I think we need was to boost
I.Q. indiscriminately.

PS: there's plenty of unexploited
technologies for steepening the
sides of the bell curve, i.e.
for reducing I.Q. differences,
at least on the low end (i.e. get
rid of most of the very low
numbers). ERASURE of differences
should not be the goal, and
will not be possible in any
case. But getting rid of a lot
of the low numbers is quite
possible.