Adaptive Scaffolding in an Intelligent Game-Based Learning Environment for Computer Science

Recent years have seen growing interest in intelligent game-based learning environments that simultaneously provide intelligent tutoring systems’ adaptive pedagogical functionalities and promote learners’ engagement through the rich narratives and situated challenges of games. A key research question posed by game-based learning is how to deliver effective adaptive scaffolding to support tailored learning processes for individual students. This paper discusses adaptive pedagogical strategies in the context of introducing computer science principles to middle school students. We explore two adaptive scaffolding strategies: (1) adaptive task selection based on students’ problem-solving performance, and (2) adaptive hint generation by exploring the problem’s solution space. Furthermore, we present specific approaches that leverage dynamic Bayesian networks and solution spaces to achieve each of these.