With the recent publication of Dan Hurley’s Smarter: The New Science of Building Brainpower, it seems timely to ponder this idea of whether you can become smarter by working through a suite of cognitive exercises.
Given my interest in cognitive systems, it sounds like something I could get behind. If we, as individuals, can increase our brainpower with some modest cognitive training, the impact on the cognitive systems in which we participate could be dramatic.
However, let’s not get carried away. A small dose of scientific skepticism might be in order.
There are commercial purveyors of brain exercises. Lumosity is one of the more visible on the web. By their account at least, they are a leader in the science of brain training. They tell us that seven published studies show the effects of Lumosity in diverse populations. The link on that page did not take me directly to those studies (which, in itself suggests something) but after a little searching, I found them here.
Not only is this research disappointing, but even its ambition is disappointing. Those who avail themselves of the brain training offered by Lumosity certainly learn something but as far as I can tell, no piece of this research tests whether people improve in their thinking, judging or deciding in anything like the situations we encounter in everyday life.
I thought it worthwhile to turn elsewhere for some insights. Jordan Lewis, writing for Slate, offers a nice summary. He cites a number of studies published in leading behavioral science journals. In brief, brain-training certainly does improve performance on the trained tasks. However, it offers no generalized benefit to overall intelligence or to other cognitive functions such as memory or attention.
Possibly, we should not take Slate as the final authority. Christopher Chabris and Daniel Simons, in their 2010 book, The Invisible Gorilla, cover this issue in their chapter 6. Their conclusion is similar. Improvement is specific to the trained task. The new skills do not transfer to other sorts of tasks.
Dan Hurley was apparently encouraged to write his book by the enthusiastic response to an article he wrote for the New York Times in 2012. In that article, he based his argument on a 2008 study by Susanne Jaeggi and Martin Buschkuehl who demonstrated a positive effect of a relatively brief amount of training on a working memory task.
Hurley offers a rather cursory account of those who are critical of this line of research but then moves quickly back to the positive stuff. He emphasizes the positive and neglects the negative, a huge body of research showing no effect of brain training on transfer to untrained tasks, even when those tasks are cognitively similar to the trained tasks. He even ignores a review by Buschkuehl and Jaeggi, his favorite researchers, which offers a more modest interpretation of this enterprise. In that review, Buschkuehl and Jaeggi observe that there is, as yet, no evidence that these training effects extend beyond laboratory tasks to standardized measures or even academic achievement or into daily life in general.
What does this say about cognitive systems?
On the one hand, it is gratifying that Dan Hurley, a journalist, uses research to develop a readable and engaging account of a scientific issue. On the other hand, it is unfortunate that he relies so heavily on questionable studies and passes so lightly over substantive arguments that cast doubt on his central claim.
The enthusiastic response to Hurley’s New York Times article shows that this is an issue that engages many people. We are all looking for a quick fix. It would be wonderful if we could make ourselves smarter with some modest cognitive training. With a balanced assessment, Hurley could have made an important contribution. Unfortunately, a balanced assessment would not sell anywhere near as well as one that promotes a quick fix.
Despite a substantial research effort, there is no evidence that brain training as discussed by Dan Hurley generalizes beyond the specific training tasks or their close cousins. I remain skeptical that this sort of training could enhance the effectiveness of the cognitive systems in which we participate in our everyday and working lives.
Worse, there is mischief about. A selectively biased review of evidence is, in itself, a cognitive system dysfunction. Many will try this and eventually be disappointed. That cannot be a good prospect for faith in evidence-based reasoning in the discussion of socially significant behavioral-science issues or indeed, issues related to any other branch of science.
Post Script: What about that style of cognitive training I discuss on the home page of my website? More on this in a future post.