Reformers, Batting Averages, and Malpractice: The Case for Caution in Value-Added Use

Abstract The essay considers two analogies that help to reveal the limitations of -value-added modeling: the first, a comparison with batting averages, shows that the model's reliability is quite limited even though year-to-year -correlation figures may seem impressive; the second, a comparison between -medical -malpractice and so-called educational malpractice, suggests that strict -accountability measures within education are out of line with legal precedent.