May 16, 2008
Fluid Intelligence is Stretchable
Improving fluid intelligence with training on working memory (Jaeggi et al. 105 (19): 6829 PNAS) presents a nice result: by running a training program for working memory fluid intelligence (as measured by standard tests) can be increased. The effect increases with the length of training, it works both for high and low performers, and the training is not on problems (in this case a version of the n-back task) similar to fluid intelligence tests (usually Raven's matrices).
This is not the first demonstration that it can be done. Torkel Klingberg, also working with computer training of working memory, demonstrated an improvement of intelligence in children with ADHD a while ago. See Training of Working Memory in Children With ADHD (Torkel Klingberg, Hans Forssberg, and Helena Westerberg, Journal of Clinical and Experimental Neuropsychology, 2002, Vol. 24, No. 6, pp. 781-791) and Computerized Training of Working Memory in Children With ADHD—A Randomized, Controlled Trial (Klingberg et al. J Am Acad Child Adolesc. Psychiatry, 44:2, feb 2005). Anecdotally his program appear to have worked in healthy people too. According to him the key trick is to make the training adaptive: the difficulty of the working memory task has to increase as people get better, or there will be no improvement. There has been some discussion about what is "really" improved, but the effects on other domains suggest that whatever it is it is useful.
Of course, we have not yet any strong evidence that this training has beneficial effects in real life situations, but that ought to be searched for once the basic effect has been replicated a bit better.
Makes me want to write a training applet to start optimizing my working memory.