This paper focuses on the advising capabilities of intelligent learning-by-doing systems. We demonstrate the importance of finding out how students use the coaching resources that ITS's make available to them. In particular, which coaching affordances do students use, and which do they ignore? To what extent do system designers' expectations about usage match actual use? We present our observations and analysis of students' use of the advising capabilities of a learning-by-doing system for electronic fault diagnosis called Sherlock II (e.g., [3] and [4]). We found that students sometimes depend upon coaching resources to solve the problem for them, and thereby miss opportunities to learn from these resources. Two main principles about the design of effective advising resources stem from this work: (1) design for learning conversations in which students play an active role, and (2) facilitate the integration of different types of knowledge (e.g., conceptual and procedural knowledge).
CITATION STYLE
Katz, S., Lesgold, A., Eggan, G., & Greenberg, L. (1996). Towards the design of more effective advisors for learning-by-doing systems. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 1086, pp. 641–649). Springer Verlag. https://doi.org/10.1007/3-540-61327-7_164
Mendeley helps you to discover research relevant for your work.