The Professor Lieury from the University of Rennes in France made an experience to see if the brain training games were really efficients. You can find below the protocole and the results observed.

'Before and after the programmes, the children were set a variety of tasks - logic tests, memorising words on a map, doing sums and interpreting symbols. Researchers found that children using the Nintendo DS system failed to show any significant improvement in memory tests. They did do 19 per cent better in mathematics - but so did the pencil-and-paper group, while the fourth group did 18 per cent better. When it came to memorising, the pencil-and-paper group recorded a 33 per cent improvement, while the Nintendo children were 17 per cent worse. In logic tests the Nintendo children registered a 10 per cent improvement, as did the pencil-and-paper group. The children who had no specific training improved 20 per cent'

The results weren't enought significants to prove that this kind of game are better than a traditionnal brain training, however it is surely better than do nothing.

(Source : Timesonline)