By Gregory Kellett, a cognitive neuroscience researcher at SFSU and UCSF, and science writer for Lumos Labs.
A study conducted by Martin Buschkuehl and Susanne Jaeggi in John Jonides’ lab at the University of Michigan indicates that it is possible to improve on measures of fluid intelligence by training one’s working memory.
The concept of fluid intelligence (gF) as defined by its founder Raymond Cattell is “…the ability to perceive relationships independent of previous specific practice or instruction concerning those relationships.” Fluid intelligence contributes to abilities like learning and problem solving. It is distinct from its counterpart, crystallized intelligence (cF) which involves “…abilities that have obviously been acquired, such as verbal and numerical ability, mechanical aptitude, social skills, and so on.”
Fluid intelligence tests usually entail completing visual patterns of some kind. Performance on such tests typically declines after reaching a peak in early adulthood. This study, however, offers evidence that it’s possible to improve fluid intelligence, at least temporarily.
The researchers used a computer-based working memory task called the “dual n-back” to simultaneously administer auditory and visual stimuli in sequence. A response was required whenever one of the presented stimuli (visual or auditory) matched a previously presented stimulus n positions back in the sequence. Four groups trained daily for either 8, 12, 17 or 19 days, with each group being matched by a control group that did not have training. Pre and post tests of fluid intelligence were given to all groups.
What the study found:
- The working memory training significantly improved performance on the fluid intelligence tests.
- Fluid intelligence performance improved in proportion to the amount of training received.
- Working memory (as measured by digit span) also improved significantly.
The authors suggest that the above effects were due primarily to an increased ability to control attention.
References:
Cattell, R. B. (1971). Abilities: Their structure, growth, and action. New York: Houghton Mifflin.
Jaeggi, S., Buschkuehl, M., Jonides, J., Perrig, J. (2008). “Improving fluid intelligence with training on working memory.” PNAS- Proceedings of the National Academy of Sciences
9 Comments
My wife and I have written an open-source/on-line implementation of dual-n-back. If you’d like to try it out, it’s at: http://www.soakyourhead.com/dual-n-back.aspx
Nicely done Erik.
Thanks for sharing that.
I noticed that as the aim is to provide the best training experience a training type module might gradually develop. This would include working up to maximum at different rates. So the general training sessions are a bit like an aerobic training a good general aerobic and flexibility training that exercises different areas and skills from minimum to maximum. When there is an average for the maximum people usually get more specific training might involve only a couple of intermediate levels to warm up to near their maximum and then one technique would be working away under their maximum for longer or another could be pushing up through their best in waves that reach progressively higher. The increments would be less near the maximum to allow people to exercise more at that level. The idea being to strengthen those skill superhighways through longer training at near the maximum level like for heart health. The strength training equivalent would seem like the total number of trucks on the superhighways and so perhaps involve something like 4 reps of 12 at a maximum. Additionally as brains also have their own unique structure and function there might be different training techniques that are unique to braintraining.
I’ve done both the program at soakyourhead.com and bought the program from the original creators at the University of Bern (for 50 EUROS!). In my opinion, the soakyourhead.com gives more feedback and is free. The “premium program” is nicely programmed, but is a lot of money (I’d be happy to share the program free with anyone doing research, if they want to contact me, and I’ll contact the University of Bern for permission to share).
Although I have an official IQ of about 130, my performance on the dual n-back task has been lacking. I can only get to the third level, and can only get about half the questions correct. I have begun a mini-study with soakyourhead.com to see if I can actually improve significantly (those in the Jaeggi et. al. study improved to level 7! within 19 days).
Here are my results:
Day 1: 573 points; average N score of 2.25
Day 2: 548 points; average N-Score of 2.15.
Anyone interested is welcome to email me. Please do share any way to improve.
Joe
Contact me at Joe_Klunder@brown.edu if you want to speak more.
Joe
Good point and very true “Integral”. As with any training, it is important to keep tasks adaptive IE Challenging yet not discouraging.
Its a fine balance.
I am first time see this blog. Amaze to find such great post for brain training soon I will become the most intellegent by reading your post.
I have done the soakyourhead game and after about a month of consistent – but far from daily – practice – like say maybe 3 full sessions a week, I have seen some progress. From an initial average n-back of 2.4 I’ve progressed to an average of 3.55. I’ve also played the “official” game from U of Bern and though my competency does seem to “transfer” to that game, I still find it harder than the soakyourhead.com game because it does not offer the same kind of in-session feedback as the soakyourhead game does – your stats/scores are only revealed after the game is over. I wonder ultimately how this affects training – whether or not the in-session feedback hurts or helps the efficacy of the training…does the work of the official creators mention anything to this effect?
Hey Bobby,
I haven’t seen any specific n-back related research on in-session feed back. I would imagine that, in general, how this might affect performance would depend on the type of game and form of feedback… IE distracting under certain conditions and motivating in others.
Post a Comment