Date of Original Version
Abstract or Table of Contents
We can reliably build “second generation” intelligent computer tutors that are approximately half as effective as human tutors. This paper evaluates two interface enhancements designed to improve the effectiveness of one successful second generation tutor, the ACT Programming Tutor. One enhancement employs animated feedback to make key data structure relationships salient. The second enhancement employs subgoal scaffolding to support students in developing simple programming plans. Both interventions were successful, but had very different impacts on student effort required to achieve mastery in the tutor environment and on subsequent posttest accuracy. These results represent a step forward in closing the gap between computer tutors and human tutors.