The market for brain-training software continues to grow, but evidence of the programs’ ability to boost memory or intelligence in a broadly applicable way (rather than simply making people better at the task they are practicing) remains scarce. New studies offer a tantalizing suggestion that certain programs may work--but the bulk of the research is murky.

Neuroscientist Peter Snyder of Brown University reviewed nearly 20 software studies and concluded that, as a group, they were underwhelming. They are marred by flaws that induce confounding factors, such as a lack of control groups and follow-up, Snyder warns. More than a third of those he reviewed were too shoddy even to include in the analysis he printed early this year in the journal Alzheimer’s & Dementia. “You’d be surprised at what gets published,” he says. Although some products claimed to treat dementia, Snyder did not find any evidence to back such claims.

The human brain benefits from taking breaks, but a computer can go strong all day. So why not enlist its power in cutting-edge research while you are away reenergizing over coffee or lunch? [More]