The Roots of Modern Brain Training through Games

In 2008, Susanne M. Jaeggi, Martin Buschkuehl, John Jonides, and Walter J. Perrig presented a paper called “Improving fluid intelligence with training on working memory” in the “Proceedings of the National Academy of Sciences of the United States of America”.

Jaeggi et al conducted experiments on fluid intelligence by having subjects play a “dual n-back” game. (FAQ at gwern.net). Players are shown a grid which contains a letter in one square. Then a different letter in a different square is shown. Then a third, and players are asked if this third letter/position is the same as the first. A fourth is shown, and players answer if it is the same as the second. And so on. The challenge is to keep new information in memory while recalling and processing an older memory. The experiment was to see if, as players improve at the game, they would see improvements in their “working memory.”

Subjects were given a preliminary IQ test, and then played the game over spans ranging from 8 to 19 days. A subsequent IQ test showed better results, and those who played the game over 19 days showed more improvement than those who played for only 8.

Subsequent experiments, (including two more led by Jaeggi, with different teams) one led by Qiu Feiyue, and one led by Susanne Schweizer have produced similar results. These results contradict the conventional wisdom that IQ is hardwired and immutable. And while research into these methods for improving intelligence s still relatively new, controversies about improving intelligence are forcing us to reconsider just what the nature of intelligence is.

Authorities such as Michael Merzenich and Robert Plomin have provided additional grist for the controversy mill. Merzenich argues that improving one’s ability at a game merely makes on better at the game itself, and doesn’t necessarily translate to improvements in other areas. Plomin points out that since the idea of intelligence being unchangeable is flawed, science that contradicts this notion is not very remarkable.

But a collection of other studies show that, if nothing else, brain stimulation does result in physical changes in the brain, whatever the current definition of “intelligence” might be. Consider a study done of London Taxi drivers, which found that after intense practice for their licensing exams, (up to four years of practice) these cabbies had enlarged areas of the hippocampus. No one’s giving them Nobel prizes, but no one can deny the intellectual achievement of memorizing 25,000 streets.

And then there’s the Nun Study at the University of Kentucky, which has found a correlation between positive emotions and low incidence of Alzheimer’s. Intriguingly, they even found cases where nuns with physical (post-mortem) evidence of disease did not display any cognitive symptoms during their lives.

Perhaps this is pushing the boundaries of what IQ means, but if a group of women are living, happily, thirty and forty years past the average age of death, I’d call that pretty smart.

New Yorker Calling Brain Training Bunk is Bunk

Hop over to the New Yorker blogs and read up on all kinds of things. The GOP, avant-garde artists, documentaries, hot tech news. Take your pick of categories, all of them smart. Smart blogs written by smart people for smart people. But even smart people get things wrong sometimes.

In “Brain Games Are Bogus” Gareth Cook asserts, unequivocally, that brain training can not make you smarter. He bases this assertion on the result of some meta-analysis done by “a pair of scientists in Europe” who gathered “all of the best research—twenty-three investigations of memory training” from around the world. According to these guys, playing games can make a person good at playing games, “but not at anything anyone might care about in real life.”

Playing games does not, according to Cook, transfer to an increased ability to do arithmetic, or “to other measures of intelligence.” He says these scientists have settled “this controversial issue.”

No, they haven’t.

The “controversial issue” is not whether games improve the mind. The controversial issue is: what exactly is intelligence? Give people many arithmetic tests and they’ll get good at arithmetic tests. Give people many IQ tests and they’ll get good at IQ tests. It’s not the processes that don’t work, it’s the conclusion drawn from the processes.

Gamify these tests, and some factors that lend themselves to learning resistance start to melt away (such as boredom). Cook doesn’t bother to mention this implicit versus explicit motivation. But one of the scientists he’s criticizing does mention motivation, suggesting that we’re only getting half the picture in this analysis of the brain training controversy. Actually, less of the picture—I have on my desktop a list of brain research citations. It’s 500 pages long. It contains a few orders of magnitude more than just 23 studies.

I’m not saying you shouldn’t read the New Yorker. Indeed, the good news is the first comment beneath this blog entry mentions an ongoing study where Catholic nuns are allowing their brains to be examined as they age. So far a direct correlation has been found between brain stimulation and brain health.

Nor am I in the least suggesting Cook is not smart. But I’d love to see what he thinks after playing some brain training games for a few weeks.