Sinking ship. A new study casts doubt on claims that computer games that use exercises like this one, which requires figuring out which object is heaviest, can make people smarter.


Do Brain-Training Programs Work?

Play a computer game, boost your IQ—that's the claim made by some software companies peddling so-called brain-training programs. It's probably an empty promise, according to the largest study to date of brain-training software, which finds no evidence of general cognitive benefits. Yet the study's limitations give brain-training advocates plenty to gripe about.

The idea for the study originated with a BBC science television show, Bang Goes the Theory. Producers contacted Adrian Owen at the MRC Cognition and Brain Sciences Unit in Cambridge, U.K., to help design an experiment to test the efficacy of computer brain training. Many of these programs are set up like a game, and playing along supposedly boosts memory, attention, and other cognitive functions. But few rigorous studies have been conducted on them, and many researchers question whether even the best programs do anything more than make people better at the game itself. For example, there's little solid research to suggest that using these programs has a beneficial effect on overall cognition that carries over into daily life.

In the new study, Owen and colleagues developed two online training programs and tested them in 11,430 healthy adults who registered on a Web site set up by the BBC. One group trained on a program that emphasized reasoning and problem-solving skills, and another group trained on a program that emphasized different skills, including short-term memory and attention. A third, control group, essentially did busywork, hunting for answers to general knowledge questions on the Internet. All participants were asked to "train" for at least 10 minutes, three times a week for 6 weeks, and all received a battery of cognitive tests before and after this 6-week period.

Not surprisingly, people in both training groups got better at the tasks they actually practiced. But that's as far as it went. "None of the brain-training tasks transferred to other mental or cognitive abilities beyond what had been specifically practiced," Owen's co-author and MRC colleague Jessica Grahn said at a press conference this morning announcing the results, which are published online today in Nature.

Brain-training advocates are not impressed, however. "This article makes big claims out of a single negative finding," says Torkel Klingberg, a neuroscientist at the Karolinska Institute in Stockholm and founder of a company that makes computer brain-training programs to improve working memory and other cognitive functions. Klingberg has published one of the few studies demonstrating that the benefits of training can generalize beyond a specific task. Among other qualms with the new study, he notes that subjects trained an average of only about 3 hours in total, and—for all the researchers know—may have done so with the TV blaring in the background or other distractions.

Although Klingberg says he agrees with the authors that many of the brain-training products on the market are probably ineffective, he argues that just because Owen and colleagues developed a training program and showed it didn't work doesn't mean that all such programs are worthless. "That does not mean that cognitive functions cannot be trained, or that all training paradigms lack effect," he says.