Does brain training really work?

Scientific experiments and their findings should be evaluated by independent experts in a process called "peer review". This is the minimum standard by which the quality of any research is judged.
Very little research on commercial brain trainers has been peer reviewed.
Instead, manufacturers often link to white papers and conference reports, or quote unsupported testimonials from "satisfied users".
None of these are acceptable substitutes for peer review.
Several companies use pictures of brains lighting up to support claims that their brain training is effective.

BRAIN IMAGING 'NOT PROOF'
Brain scans and other imaging may look good, but what do they really show?
It is a very persuasive technique, but in reality only shows how hard the brain is working, rather than how effective the training is.
Dr Jessica Grahn, a neuro-imaging expert at the Medical Research Council Cognition and Brain Sciences Unit, said: "What we are looking at with these brain scans is simply a measure of the energy that the brain is using while performing the tests.
"It does not constitute evidence that the brain is being trained, or indeed altered, in any way."

CONTROL GROUPS
A proper "control group" is the backbone of any valid scientific study.
Without a control group, it is hard to tell whether results are down to brain training, or simply because the participant is doing something - anything - to stimulate their brain.
With a brain training study, the control group of participants should do the same amount of a mental activity as the brain-training group, but using a "non-brain training" activity such as a crossword or Sudoku.
Scientists should then be able to compare the results from the two groups to see how genuinely effective the brain training tasks were.

Studies that have been used to support the claims made by commercial brain trainers often have no control group at all, or a control group that is not as mentally active as the trained group.

THE BENCHMARKING TEST
In order to see how much you have improved, you need to know where you were at the start. To do this, scientists use a "benchmarking test" - the same test, taken at the beginning and end of an experiment.

Studies used to support the claims made by commercial brain trainers often use tests for benchmarking that are identical - or very similar - to the tests used during training.
Research has suggested Tetris can help reduce traumatic stress.

But all this proves is that practice makes perfect: if you played Tetris for a month and got better at it, should we conclude that playing Tetris is a useful form of brain training? Probably not.
Benchmarking tests should be different from the brain training tasks. This way, scientists can see if the effects of brain training are transferable to other mental tasks. If you played Tetris for a month and found, say, your memory was suddenly much better, it would indicate that playing Tetris has improved your brain in some way other than making you good at Tetris.

WHAT WORKS FOR ONE MIGHT NOT WORK FOR ANOTHER
Does the research show the brain trainer is effective in participants who are similar to you?
Just because a system claims to improve the memory of a group of elderly participants, for example, it does not necessarily mean that it will have the same - or indeed any - effect in younger users. The BBC's Brain Test Britain experiment, created with the Medical Research Council's Cognition and Brain Sciences Unit and the Alzheimer's Society, has made sure it meets all the accepted standards for a scientifically valid study.

By Dr Adrian Owen Medical Research Council, Cognition and Brain Sciences Unit