When Does Collaborative Interaction Support Learning of Computational Thinking Among Undergraduate Students

Collaborative learning has proven to be a productive approach across domains including computer science. Especially for novices, collaboration among learners with diverse experiences, values, and knowledge can be especially effective. However, there is limited research on how collaborative interactions among learners from different undergraduate disciplines manifest during the learning of computational concepts. This qualitative study investigates interaction among learners in small interdisciplinary student groups in an undergraduate computational thinking (CT) class. In analyzing the social interactions of cohort members, the Differentiating Overt Learning Activities (DOLA) framework developed by Chi was used as a starting point. Several themes (e.g. explain-to-other, self-monitor, connect/plan) suggested by Chi’s overt activities framework emerged naturally in the observations. However, several categories that were not explicitly stated in the DOLA framework also emerged. These are: asking a shallow question versus asking an elaborate question, socializing, being confused or expressing frustration, self-talk, and struggle. Analysis of interactions between cohort members revealed that most of the group interactions were active in nature. Barriers to collaborative learning were accentuated when peers were novices. Students who lacked knowledge of computational concepts found it hard to contribute to a discussion, hindering learning, whereas students who had better conceptual understanding were constructive in their interactions, elaborated on a topic, made connections with previous problems, and displayed self-awareness about their understanding.