Learning programming is becoming increasingly popular but is also very challenging. Automated tools can support students in learning programming. Tutors that support the stepwise implementation of programs through various strategies and that are easily adaptable by teachers are scarce.
In this research we develop technical concepts and solutions for constructing strategy-based and teacher-adaptable tutoring systems for programming. The novelty of our approach is the use of compiler technology to analyse incomplete student solutions to derive feedback and hints. The research will be evaluated by user data analysis of a tutor prototype and a feedback benchmark that we will design ourselves.
Supported by: NWO Lerarenbeurs
Participating CS members: Hieke Keuning, Bastiaan Heeren