nAble Adaptive Scaffolding Agent – Intelligent Support for Novices

Scaffolding techniques allow human instructors to support novice learners in critical early stages, and to remove that support as expertise grows. This paper describes nAble, an adaptive scaffolding agent designed to guide new users through the use of an analytic software tool in the ´nSpace Sandbox´ for visual sense-making. nAble adapts the interface and instructional content based on user expertise, learning style and subtask. Bayesian Networks and Hidden Markov task models provide the agent reasoning engine. An experiment was conducted in which participants were provided with one of: an adaptive scaffold, an indexed help file or a human guide. Users of the adaptive scaffold outperformed users of the indexed help and more quickly converged with the performance of users with the human guide.