Fostering the Intelligent Novice: Learning From Errors With Metacognitive Tutoring

This article explores 2 important aspects of metacognition: (a) how students monitor their ongoing performance to detect and correct errors and (b) how students reflect on those errors to learn from them. Although many instructional theories have advocated providing students with immediate feedback on errors, some researchers have argued that immediate feedback eliminates the opportunity for students to practice monitoring for and learning from errors. Thus, they advocate delayed feedback. This article provides evidence that this line of reason is flawed and suggests that rather than focusing on the timing of feedback, instructional designers should focus on the "model of desired performance" with respect to which feedback is provided. Instead of delaying feedback, we suggest broadening our model of correct behavior or desired performance to include some kinds of incorrect, but reasonable behaviors. This article explores the effects of providing feedback on the basis of a so-called intelligent novice cognitive model. A system based on an intelligent novice model allows students to make certain reasonable errors, and provides guidance through the exercise of error detection and correction skills. There are two pedagogical motivations for feedback based on an intelligent novice model. First, jobs today not only require ready-made expertise for dealing with known problems, but also intelligence to address novel situation where nominal experts are thrown back to the state of a novice. Second, the opportunity to reason about the causes and consequences of errors may allow students to form a better model of the behavior of domain operators. Results show that students receiving intelligent novice feedback acquire a deeper conceptual understanding of domain principles and demonstrate better transfer and retention of skills over time.