The intelligent tutoring system LARGO allows law students to annotate a transcript of an oral argument using diagrams which can be linked to text portions. LARGO analyzes diagrams and gives feedback to support students’ reflection. The feedback mechanisms rely on certain hypotheses about the interaction between the student and the system. In particular, a linear working mode (from beginning to end of the transcript) and a consistent and correct linking of diagram elements to the text are assumed. Based on an empirical study, this paper argues that the design of LARGO is functional and that the central interaction hypotheses are confirmed.
Niels Pinkwart, Collin Lynch, Kevin D. Ashley, Vin