Good Issues and bad tidying: what GitHub can tell us about agency in project-based group modelling work for higher education

Abstract

Collaborative project work in technology-enabled environments at university is essential for learners to become ready for an increasingly global, complex, and virtualised workplace. Research on effective pedagogical and technical design for computer supported collaborative learning in higher education (CSCL) has often taken place in synchronous contexts, using specialised technology platforms. However, large-scale changes to work and education resulting from the COVID-19 pandemic necessitate the development of pedagogical and research approaches that support students working asynchronously, in distributed teams, using collaboration platforms that extend beyond institutional infrastructure. Within the field of CSCL, knowledge building research has shown collaboration to be a complex systems phenomenon, involving the intersection of individual and collective efforts to actively advance the group’s shared knowledge, but studies analysing interaction data have been resource-intensive to conduct. Contemporary workplace platforms such as professional knowledge environments have multiple design affordances consistent with knowledge building principles, as well as the capacity to generate rich data about user activity. However, we have little understanding to date as to how these environments can support knowledge building pedagogies and facilitate associated research. This study uses a case study approach and thematic analysis to investigate the activity of three university groups engaged in a collaborative modelling task over time. It investigates how agency emerges during project work in professional knowledge environments, and how the system interaction data can extend our understanding of effective collaboration processes. The results show that the GitHub platform can support knowledge building pedagogical designs in facilitating individual and collective agency in higher education group work, and provide insights into epistemic, regulative and relational aspects of learner behaviour at individual and group levels. These findings extend our understanding of effective learning design to novel environments of a type likely to be used by our students in the workplace, and make design and methodological contributions to research on computer-supported collaborative learning.