Gritty sand in the gears of "reform"

Every so often the New York Times Education section publishes something interesting.  Most recently, they published an op-ed by Angela Lee Duckworth, in which she argued that "grit" (a psychological concept that she has been a leader in studying) should not be measured in the classroom with the goal of attaching high stakes.  As soon as something goes from the lab to the world of high stakes, all sorts of other behaviors come into play.  Kudos to her for recognizing this.  She's making the same point that I made in my responses to Lani Guinier's Tyranny of the Meritocracy:  As soon as you go beyond This One Special Situation to a world of competition and financial incentives, people will try to game your lovely research construct.