posted on 1987-01-01, 00:00authored byJohn R.(John Robert) Anderson, Artificial Intelligence and Psychology Project.
Abstract: "This paper starts with a sketch of the basic learning processes that we think are at work in the subjects we have tutored. This basically involves students doing a causal analysis of the examples and trying to extend that analysis analogically. The most straight forward ways errors can occur in this analogy process is for students to choose an inappropriate example to map to the current solution. This is certainly not the only way misunderstandings can arise and it is of interest to get a sense of what fraction of student errors may be explained in this fashion. Therefore, I will present a list of some of the dominant misconceptions that students have displayed interacting with our tutors and see how many can be interpreted in this fashion.Then, I will speculate on how many of the subtraction errors that were the focus of Brown and VanLehn's analysis can be so explained. Finally, I will compare this analysis of the source of errors with the analysis offered by VanLehn (1983, 1986)."