#348: ClassMosaic: Leveraging Solution Diversity to Teach Computational Problem-Solving

A key challenge in K-12 Computer Science (CS) education is teaching computational Problem-Solving (PS). PS instruction research emphasizes the importance of structured whole- class discussions to reflect on solution strategies to foster PS competences. However, orchestrating whole-class discussions is challenging for teachers. As such, we drew on PS- and dashboard research to design ClassMosaic, a classroom orchestration tool that supports productive PS-teaching practices. ClassMosaic provides real-time insights into student work to facilitate discussions and promote collaborative learning by enabling collective comparison and reflection. We present ClassMosaic’s design principles and the first user study focusing on perceived pedagogical value with 9 CS teachers. Results show that ClassMosaic aligns with existing CPS instruction practices, and fosters new CPS-teaching strategies, including heterogeneous strategy-based grouping and structured post-discussion adaptations. Future work will further automate PS-based code clustering, and adapt ClassMosaic’s UX/UI to reduce teacher cognitive load, before in-class experiments.