Does Brain Training Really Work?

While many of us are enthusiastic about computer-based brain training, studies of how well it works in the real world have been uninspiring. It’s  possible to train people to do better on cognitive tasks, but it’s not clear that the training carries over into the real world. Does brain training really work? A new study of more than 10,000 people says: Maybe not.

The study, reported in the journal Nature, highlights the problem called generalization. Most studies of computer-based brain training have shown, for example, that people can readily learn to do better on the tasks on which they train, People who work on learning how to remember a list of words will usually get better at it. If you train at solving mazes, you will probably get better at solving a maze puzzle.

What’s not clear, though, is that the training you do will help you on other things. You can learn a list of words and that may help you memorize a grocery list, but all that work probably won’t help you remember where you put your keys, or help you do better on a test for your job. In the same way, you may get pretty good at solving a maze, but it not likely that will help you be a better driver.

In the study reported in Nature, viewers of a popular BBC television show names “Bang Goes the Theory” participated in six weeks of online brain training. Participants completed baseline measure of reasoning, working memory, and paired associates learning and were randomly assigned to different training groups. Participants were aged 18 to 60 years, and 11,430 completed enough of the training to contribute data.

What did the study show? Consistent with a great deal of other research, the study showed that people who trained got better on the tasks they trained on, but there was essentially no evidence that they got better on the baseline tasks. The authors of the study give an illustration: people in the memory training group improved in their ability to remember numbers by about 3/100 (three hundredths) of a digit. They suggest that it would take almost four years of training to remember just one extra digit. Further, the control group in the study got better by 2/100 (two hundredths), and they didn’t even get memory training.

There are a number of legitimate criticisms of the study.Researchers could argue about the selection of tests, how participants were trained, and the small amount of training completed by some of the participants. Including a large age range may have masked improvements in some groups, such as older adults.

As always, more research is needed. But this study shows that at least short-term computer-based brain training may not make a big difference in daily life.

Reference:

Owen AM et al. Putting brain training to the test. Nature near-final version, 20 April 2010 (doi:10.1038/nature09042). Click here to go to the report.

Brain Fitness Tip: Dual n-Back

One of the most interesting findings in the field of computer-based cognitive training is based on a study by Susanne Jaeggi and her colleagues that showed that a specific kind of mental exercise can improve fluid reasoning ability. This finding is important and exciting for several reasons. One is the effect of a very small amount of training (as little as 20 minutes a day) on what many people think is a basic mental ability. The other is the idea that any kind of basic ability can be improved.

The computer-based training involves a procedure called n-back training. It’s been used for a long time in neuropsychological or cognitive assessment activities as a way to evaluate working memory. In single n-back, you have to remember, for example, the position of an object that appears intermittently in various positions on a computer screen. If you’re doing the task, you have to press a key on the computer when the object appears in the same place. In single n-back, you would watch to see if the object appears in the same two times in a row. If you are doing 2 n-back, then the object has to appear once in a specific place, then can appear somewhere else on the screen, and then appears again in the first position. The number refers to how many positions back you have to keep track of. It can go up to as many as 6 back.

In dual n-back, you have to do two n-back tasks at the same time. The other one can be auditory, for example. One task asks you to listen to numbers played on the computer’s speakers, with the same basic task. You have to remember whether a number you hear is the same as the one you just heard, or heard before the last one, and so on. It’s easier to understand if you try it out.

You can try it out in several places for free. It can be pretty challenging, but the original study showed that how well a person did the task didn’t matter so much as that the person did the task at the level that worked for them. Doing the task is a great way of training your attention, and another study has shown that this kind of training can make changes in the brain’s chemistry.

You can find try the task out at Soak Your Head (this site requires a browser plug-in called Silverlight, so you may see a message about installing it) and you can download a version for your own computer (free) at  Brain Workshop. Another free online version is at The Mindflow.