Children's Interactions with Inspectable and Negotiated Learner Models

The Learner Model of an Intelligent Tutoring System (ITS) may be made visible (opened) to its users. An Open Learner Model (OLM) may also become a learning resource in its own right, independently of an ITS. OLMs offer potential for learner reflection and support to metacognitive skills such as self-assessment, in addition to improving learner model accuracy. This paper describes an evaluation of an inspectable and a negotiated OLM (one that can be jointly maintained through student-system discussion) in terms of facilitating self-assessment accuracy and modification of model contents. Both inspectable and negotiated models offered significant support to users in increasing the accuracy of self-assessments, and reducing the number and magnitude of discrepancies between system and user beliefs about the user’s knowledge. Negotiation of the model demonstrated further significant improvements.