http://educationnext.org/grinding-the-antitesting-ax/ Eric Hanushek claims the National Research Council's report about the incentives of test-based accountability was more bias than evidence. At the core of this argument is 0.08 standard deviations. NRC says that's 3 percentile points. Hanushek says it represents a return on investment of 9,189 percent. I think I know what Mark Twain would have said about Hanushek's figure.