Covering Possible Reasoning Errors for Intelligent Tutoring Systems: Order of Expression Evaluation Case

Abstract

One of the problems in computer-supported education is defining meaningful errors that students can make during a complex automated exercise. A precise set, describing what can go wrong in the student’s reasoning, allows better measurement of students’ knowledge, asking pointed follow-up questions to stimulate the student’s thinking, and providing precise explanatory feedback. We describe a method for building a set of possible low-level errors in reasoning during solving a complex task. We demonstrate an example of using this method for the task of determining the order of evaluation of a programming language expression and discuss prospects of applying the described method.