Sep 22, 2010

Technology in Education - More questions than answers

by Laura Starita

It is that “back to school” time again and with it a surge in thinking about the best ways to improve education outcomes. The New York Times magazine chimed in this weekend with a long piece by Sara Corbett about an experimental school in Manhattan that is built around the concept of using video games as learning tools. I don’t mean there is one class about gaming in the curriculum. Instead, every class uses gaming and other computer-based learning methods to communicate the core curriculum.

Both Tim and I have written about the love affair in philanthropic circles with investing in technological interventions that are at best unproven, even in sectors where there are known programs that work.  Computer-based learning is one those interventions that seems perennially popular despite high expense and few results. Games and software that “teach” lessons seem like an incredibly attractive options for (at least) two reasons: They can be used to teach and thus augment the instruction done by teachers, especially in environments where teachers are few and suffer from low incentives; and even if students’ knowledge and skills do not improve faster than their non-techy peers, they still gain computer literacy skills that will raise their marketability in service and information economies. And yet there is no evidence that computer-based learning actually achieves that promise. The American fascination with Baby Einstein-like education tools has come to nothing, and at least one developing-world study conducted in India showed that minimally-trained part-time tutors were as effective at improving student math scores at lower cost than a computer-based learning program on the same material. (For a good summary of existing research on computer-based learning tools, see Tim’s Miller McCune article here).

And yet, Corbett’s article made me think—if not rethink—some of these views. The student’s in the school Corbett profiled do not show improved test scores over their peers in Manhattan, but that might not mean much. Our current tests, after all, are designed from our current system. More interesting was the discussion about the way kids learn when they play games. Success comes when they get the highest level, and no gamer expects to get there without having their avatar die many times before they get there. Failure, in short, is an expected and ingrained part of success. The key part is that the game will let you fail as much as you need to—it will always give you another chance. The point, I think, is not that we learn from failure—a questionable platitude—but that we (can) learn from persistence. 

Sports is the closest analogy to this kind of learning, since it also requires players to fail a lot on the way to success. Failures come small (you lose the ball to an opponent), or large (you lose the tournament). And you get knocked around a lot, which is only fun until it isn’t. I could not have been much older than seven when I realized that if I fell down and stayed down my soccer coach would pull me out of the game. I wanted to play, so I got up. That kind of persistence can’t really be taught. It’s more akin to hunger than any learned behavior I can think of. Most parents, teachers, and child advocates wants their kids to be hungry in that way—to want to intellectually get up, even after a dozen falls.  If philanthropy can find out how computers might serve that aim, than the investments in gaming experiments like the one in Manhattan will more than pay off. But let’s be thoughtful. Don’t forget to fund the less sexy, human-based programs we know will work in favor of technological (and potentially empty) allure.