The Invention Lab: Using a Hybrid of Model Tracing and Constraint-Based Modeling to Offer Intelligent Support in Inquiry Environments

Ido Roll, Vincent Aleven, Kenneth R. Koedinger

Exploratory Learning Environments (ELE) facilitate scientific inquiry tasks in which learners attempt to develop or uncover underlying scientific or mathematical models. Unlike step-based Intelligent Tutoring Systems (ITS), and due to task characteristics and pedagogical philosophy, ELE offer little support at the domain level. Lacking adequate support, ELE often fail to deliver on their promise. We describe the Invention Lab, a system that combines the benefits of ELE and ITS by offering adaptive support in a relatively unconstrained environment. The Invention Lab combines modeling techniques to assess students’ knowledge at the domain and inquiry levels. The system uses this information to design new tasks in real time, thus adapting to students’ needs while maintaining critical features of the inquiry process. Data from an in-class evaluation study illustrates how the Invention Lab helps students develop sophisticated mathematical models and improve their scientific inquiry behavior. Implications for intelligent support in ELE are discussed.

The final publication is available at Springer via https://doi.org/10.1007/978-3-642-13388-6_16.