The results from using incentives are mixed. In some cases, incentives have been very cost effective: paying elementary school children in Dallas $2 for each book they read leads to substantial test score gains. On the other hand, a number of other programs aimed at older kids have been less effective. A lot of things change across the various experiments, but one hypothesis Roland puts forth in his academic paper is that better results will be obtained when focusing on inputs that the student can directly control (e.g. turning in homework, showing up for school, wearing a uniform), instead of outcomes (test scores, grades, etc.).I imagine part of the reason is that students haven't learned the connection between inputs --> outputs. Giving them a goal of improving outputs doesn't help them discover that connection. We'd like to think it offers enough of a motivation that the discover it on their own, but clearly that is one of the important parts of learning, and it needs to be taught and encouraged.
Thursday, January 13, 2011
Economists are fond of saying incentives matter. We believe grades are incentives, yet they often fail to achieve the ends we seek. Alfie Kohn is famous for his book Punished by Rewards in which he advocates the abolition of grades. But it always struck me as a case of throwing out the baby with the bath water. Just because a set of incentives does not achieved the desired outcomes, doesn't mean the incentives are intrinsically bad, they are possibly just misaligned. Roland Fryer has been finding out how difficult it is do properly align them.