The right sort of video game can increase your intelligence

June, 2011

Games that use the n-back task, designed to challenge working memory, may improve fluid intelligence, but only if the games are at the right level of difficulty for the individual.

It has been difficult to train individuals in such a way that they improve in general skills rather than the specific ones used in training. However, recently some success has been achieved using what is called an “n-back” task, a task that involves presenting a series of visual and/or auditory cues to a subject and asking the subject to respond if that cue has occurred, to start with, one time back. If the subject scores well, the number of times back is increased each round.

In the latest study, 62 elementary and middle school children completed a month of training on a computer program, five times a week, for 15 minutes at a time. While the active control group trained on a knowledge and vocabulary-based task, the experimental group was given a demanding spatial task in which they were presented with a sequence of images at one of six locations, one at a time, at a rate of 3s. The child had to press one key whenever the current image was at the same location as the one n items back in the series, and another key if it wasn’t. Both tasks employed themed graphics to make the task more appealing and game-like.

How far back the child needed to remember depended on their performance — if they were struggling, n would be decreased; if they were meeting the challenge, n would be increased.

Although the experimental and active control groups showed little difference on abstract reasoning tasks (reflecting fluid intelligence) at the end of the training, when the experimental group was divided into two subgroups on the basis of training gain, the story was different. Those who showed substantial improvement on the training task over the month were significantly better than the others, on the abstract reasoning task. Moreover, this improvement was maintained at follow-up testing three months later.

The key to success seems to be whether or not the games hit the “sweet spot” for the individual — fun and challenging, but not so challenging as to be frustrating. Those who showed the least improvement rated the game as more difficult, while those who improved the most found it challenging but not overwhelming.

You can try this task yourself at


Jaeggi, Susanne M, Martin Buschkuehl, John Jonides, and Priti Shah. “Short- and long-term benefits of cognitive training.” Proceedings of the National Academy of Sciences of the United States of America 2011 (June 13, 2011): 2-7.

[1183] Jaeggi SM, Buschkuehl M, Jonides J, Perrig WJ. From the Cover: Improving fluid intelligence with training on working memory. Proceedings of the National Academy of Sciences [Internet]. 2008 ;105(19):6829 - 6833. Available from:


Is this a cause or just a correlation?

Could this data set not also mean that the children who improved the most were the best at the fluid intelligence tests to begin with?

Re: Is this a cause or just a correlation?

If anything, it's more likely to be the other way around. It is possible that those who didn't improve much were already at their own 'ceiling'. Those who improved the most might be the ones with the lowest fluid intelligence -- because they had more room for improvement. However, the researchers did check for this. They didn't find any significant group differences between those children with high fluid intelligence at the beginning of the study and those with low intelligence, as far as the amount of transfer (to the abstract reasoning task) is concerned. Nor was there any correlation between the gain in fluid intelligence and performance on the first two training sessions. So while that might be a minor factor, it doesn't seem to be the critical factor.