If you train intensively at long-distance running, you’ll find it easier to climb stairs or ride a bike. The running will boost your aerobic fitness and strengthen your leg muscles, providing benefits that transfer to other activities. Does our brain work in the same way? If we train ourselves on a specific mental task, do we become sharper across the board?
There’s a multi-million dollar industry that would like you to believe the answer is yes. Best-selling “brain training” games like Brain Age purport to give your brain a “the workout it needs” through a combination of word puzzles, number problems, Su Doku and more. Unfortunately, there is little good evidence that these games improve anything beyond performance on a specific task.
There are exceptions. Susanne Jaeggi from the University of Michigan has found that a simple exercise called an “n-back task” could increase the “fluid intelligence” of elementary and middle-school children (well, some of them; more on this later). It’s the latest of a series of studies showing that practicing at a single task can lead to a broader intellectual boost.
Fluid intelligence is a broad concept that includes abilities like abstract reasoning, solving new problems, spotting patterns and drawing inferences, rather than relying on knowledge, skills or experience. The n-back task isn’t meant to train all of these. Instead, it is meant to improve working memory – the ability to hold and manipulate pieces of information in our head. A good working memory is essential for problem-solving and reasoning, so Jaeggi reasoned that training the former would improve the latter, in the same way that building a runner’s fitness improves their cycling.
She was right. In 2008, she showed that students who trained on the n-back task did much better on fluid intelligence tests than peers who didn’t receive any training. The more they trained, the better they became. It was an exciting result. Fluid intelligence is thought to be strongly influenced by genetics and unaffected by knowledge and education. The prospect of improving it through a simple task is an enticing one.
Now, she has turned her attention to children. She recruited 62 children, aged between seven and ten. While half of them simply learned some basic general knowledge questions, the other half trained with a cheerful computerised n-back task. They saw a stream of images where a target object appeared in one of six locations – say, a frog in a lily pond. They had to press a button if the frog was in the same place as it was two images ago, forcing them to store a continuously updated stream of images in their minds. If the children got better at the task, this gap increased so they had to keep more images in their heads. If they struggled, the gap was shortened.
Before and after the training sessions, all the children did two reasoning tests designed to measure their fluid intelligence. At first, the results looked disappointing. On average, the n-back children didn’t become any better at these tests than their peers who studied the knowledge questions. But according to Jaeggi, that’s because some of them didn’t take to the training.
When she divided the children according to how much they improved at the n-back task, she saw that those who showed the most progress also improved in fluid intelligence. The others did not. Best of all, these benefits lasted for 3 months after the training. That’s a first for this type of study, although Jaeggi herself says that the effect is “not robust.” Over this time period, all the children showed improvements in their fluid intelligence, “probably [as] a result of the natural course of development”.
These results are more qualified than those from the first experiment in 2008. In that one, the difference between the n-back group and the controls was very clear. Here, only some of the children who trained their working memory developed better fluid intelligence as a result. And Philip Ackerman, who studies learning and brain training at the University of Illinois, says, “I am concerned about the small sample, especially after splitting the groups on the basis of their performance improvements.” He has a point – the group that showed big improvements in the n-back training only included 18 children.
Robert Sternberg, an intelligence researcher at Oklahoma State University, is more convinced. He says, “This paper adds weight to the evidence that fluid ability is trainable. We had known from the so-called Flynn effect that fluid ability has been increasing at least since the early 1900s. But it was not as clear that fluid ability could be increased within a generation. This study suggests that indeed it can be – that this important aspect of intelligence is not static and fixed, but rather is dynamic and flexible.”
Why did some of the children benefit from the training while others did not? Perhaps they were simply uninterested in the task, no matter how colourfully it was dressed up with storks and vampires. In Jaeggi’s earlier study with adults, every volunteer signed up themselves and were “intrinsically motivated to participate and train.” By contrast, the kids in this latest study were signed up by their parents and teachers, and some might only have continued because they were told to do so.
It’s also possible that the changing difficulty of the game was frustrating for some of the children. Jaeggi says, “The children who did not benefit from the training found the working memory intervention too effortful and difficult, were easily frustrated, and became disengaged. This makes sense when you think of physical training – if you don’t try and really run and just walk instead, you won’t improve your cardiovascular fitness.” Indeed, a recent study on IQ testing which found that they reflect motivation as well as intelligence.
Jaeggi’s next goal is to find out what makes n-back training effective, whether only specific groups of people will benefit from it, and what conditions lead to the best results. “We are quite at the beginning when it comes to that. We know that training time is an issue (the more training, the better), we know that improvement in training is a factor, but other than that, we don’t know much. There is some evidence showing that distributed training is better than massed training. There are hundreds of other issues that wait to be clarified.”
For the moment, it seems that the flood of n-back apps that followed from Jaeggi’s 2008 study were premature. The task shows some promise, but it isn’t an all-purpose route to a higher intellect. As ever, things are more complicated than that.
Sternberg also wants to see if working memory training can improve performance at “more serious endeavours” like school work or job performance, instead of simply on puzzle-like tests. He doesn’t want the n-back tasks to go the way of IQ test – something that is meant to predict a person’s performance, but has become a criterion in itself. “Parents and even schools sometimes seem more concerned with test scores than with the achievement they are supposed to predict or to measure,” he says. In the case of the n-back studies, it would be a shame if people took their scores to mean an automatic intelligence boost.
Finally, Sternberg says, “Fluid ability, although important, is only part of intellectual abilities broadly defined. Other abilities, such as creative thinking, practical thinking, and wisdom-based thinking, are at least as important. In our world, we seem more to lack wisdom and ethics as applied to everyday situations than we do the abstract-thinking skills measured by fluid-ability tests.”
Reference: Jaeggi, Buschkuehl, Jonides & Shah. 2011. Short- and long-term benefits of cognitive training.
For more on the new study, see Jonah Lehrer’s take
More on brain-training and IQ:
June 13th, 2011 at 4:34 pm
It certainly sounds like correlation rather than causation. The kids who did best at an intelligence-related task also did best at some other intelligence-related task later on (what exactly IS a test of fluid intelligence? I thought quantifying intelligence was problematic).
June 13th, 2011 at 6:00 pm
1) Intelligence does exist, it can be measured, it does make a difference.
Department of Education schooling is useless, kills inspiration and curiosity, is mind-numbingly tedious, makes no connections to anything, and is forgotten immediately after the test.
2) The Gifted are more intelligent than the mediocre.
3) The Gifted can be identified in very early childhood.
4) The Gifted can be ruined but the mediocre cannot be elevated.
5) The Gifted given opportunity make more of it than the mediocre do.
6) The Gifted get smarter with feeding, the mediocre get fatter.
7) Desirable futures require Gifted architects.
9) Rather than foster brilliance we allocate for its suppression.
10) DIVERSITY is social advocacy demanding the bottom 10% be the bottom 90%. The top decile is medicated into compliance by the|rapists.
http://www.diversity.umich.edu/programs/
90+ diversity programs, zero Gifted programs.
June 13th, 2011 at 6:16 pm
I really hate coming out and saying things like this, but… Esme, you are an idiot. And completely off-the-mark in regards to human ability and potential to achieve. The mediocre can’t be elevated? What are you on about? Have you ever been in the practice of teaching, the “mediocre”, or the disabled, or the gifted? Well, I have, and I can tell you that you are 100% wrong. Also, you are doing no service to the gifted community by making such discriminatory and elitist (not to mention delusional) statements.
However, I do agree that the current educational system blows for the majority of the population.
Also, how do you define “gifted”?
June 13th, 2011 at 6:36 pm
Esme, your comment makes it clear that you are not at the gifted end.
Now what?
June 13th, 2011 at 8:15 pm
Did the author control for comfort using a computer or video game interface? Perhaps a validation test could be performed to make sure students don’t start with abnormally low scores.
I think the author could have done a better job controlling for socioeconomic status. They said in the materials and methods students were selected from Ann Arbor and Detroit school systems. There are probably high and low performing schools in both school systems. How do we know she didn’t pick the two best from either school system?
Do we know much about the heritability of fluid intelligence? Does fluid intelligence correlate with socioeconomic status in adults?
Thanks Ed.
June 13th, 2011 at 10:29 pm
Measuring intelligence will always be problematic because any test will be a projection of a hyperdimensional space (our ability to think) onto a smaller dimensional hyperplane (the test) for comparison. The angle of the ruler can greatly influence the measurement. This is a tough issue to dissect.
June 14th, 2011 at 1:17 am
Learning and playing contract bridge will exercise the working memory quite a bit. Many of us find it interesting and enjoyable. Kids as young as 8 can learn to play reasonably well. There is a lot of use of memory in the process of playing, remembering bids and cards played. To play at all well one must reason from the limited information available. Of course, no big money corporate interest will make money from this.
June 14th, 2011 at 8:10 am
More info on actual practice of this exercise here
June 17th, 2011 at 4:30 am
@Io, there’s definitely something there, but the point was that the students who improved at the task showed the effect, not the ones who did the best. I’d hope there’s some effort in the study to do a comparison between improvement on the test and performance on the test, to separate these factors out, but it sounds like based on the experimental design this may not be the case.
@Robert, I learned about probability through card games, thanks to a math teacher who enjoyed them. Still recall most of it, which is more than I can say for some subjects. It’s interesting that known memory- and logic-reliant tasks that have been proven to keep people’s attention, like bridge, aren’t more well-studied in this case. I suppose it makes sense, though, since the games are complex, and at this point researchers are trying to ferret out what elements are contributing.
June 20th, 2011 at 4:11 am
@Zimmerman: That is a valid point. But there could be alternative ways of looking at the challenge of projecting the hyperdimensional space on to the smaller plane of tests. The question to ask would be, what kind of intelligence results in improvement of what kinds of activities, which is what I think the direction in which the research is heading. Issues of measurement might still prevail, but test results have always been more suggestive than conclusive.
July 5th, 2011 at 11:27 pm
For operational memory, I like Sherlock, a 20-year-old game that runs in DOS. You use clues to arrange 6 items in each of 6 rows (for 6!^6 possible solutions). Each of the 36 spaces shows miniature versions of the 6 things that could go there. You can left-click on one to enlarge it and eliminate the others, or right-click on it to eliminate just the one. The working-memory training happens when you try to solve the puzzles without right-clicking, putting in only positive information. And you get to make your own pictures.
My 3-year-old niece, on the other hand, needs to work at being able to count up to, but no further than, a number she was just told. (This arose while playing a board game where her mother let her roll the die and move her pawns.)
July 7th, 2011 at 12:07 pm
@Chris, when children are at age 9-10 like the children using the game, their IQ scores tend to increase noticibly as the natural course of maturation. So it could be that children experiencing a growth spurt in IQ at the time of the training could have in turn improved on n-back, giving the illusion of more dedicated children improving iq with n-back, when really it could be children experiencing a growth spurt can therefore improve on n-back. Just my $.02.
May 8th, 2012 at 6:15 am
Will any of this replicate? That is the key…