Using Decomposed Error for Reproducing Implicit Understanding of Algorithms | MIT Press Journals & Magazine | IEEE Xplore