Learning to program is challenging, especially for novices. Self-regulation of learning, recognised as a fundamental skill for success in programming, has been extensively studied in previous research, with an emphasis on individuals' metacognitive and motivational regulation factors.
Introductory programming courses in university classrooms comprise a social setting where learning does not occur in isolation. Students’ learning regulation in classrooms is often driven by spontaneous collaborative interactions between peers, which facilitate synergy and also influence their learning motivation. Therefore, programming learning regulation in the classroom is a socially-situated phenomenon.
Several researches have explored collaborative interventions and methods for effective learning in programming. However, previous research has not addressed the role of novices’ social modes of regulation in programming and its subsequent impact on learning motivation---which is inevitable in collaborative learning settings, such as classrooms.
Thus, this thesis seeks to investigate how an individual’s programming learning behaviour unfolds in the social learning context of university classrooms. To do this, I examine students' learning behaviour using the contemporary education psychology theory of Social Modes of Learning Regulation, which explains how learning regulation emerges through social interactions and environmental contexts. I also investigate novices’ programming learning motivation using a human motivation theory known as the Self-Determination Theory, which focuses on enabling individuals to take self-directed actions, drive their own regulatory processes, and foster intrinsic motivation by fulfilling their basic psychological needs. Based on these two theories, collaborative extensions were designed for the Jupyter platform used in introductory programming courses in many universities. The extensions evolved over four years across classroom formats, including online, hybrid and in-person formats. By collecting quantitative and qualitative data, through three longitudinal studies across the three respective classroom modalities, the designs were refined and tested the effect of different interventions in supporting students' social regulation and motivation to learn programming.
Specifically, in a first study, an intervention for online introductory programming classrooms was designed to support remote learners' regulation at individual and social levels and to foster their intrinsic motivation to program. The design included affordances for individual metacognitive reflection and opportunities to progress in learning through peer interaction, enabling a type of social regulation known as 'co-regulation'. Analysis of students' behaviour using the intervention revealed sustained co-regulatory practices, such as spontaneous discussion on programming challenges and feedback on code, errors, and outputs, which helped students regulate their own learning and influence their peers' regulation. The evaluation further demonstrated that use of the intervention’s features influenced students’ SDT motivational factors, fostered co-regulatory behaviour, and sustained their intrinsic motivation for programming.
Based on the findings of the first study, a new intervention was developed for a hybrid-taught introductory programming classroom to foster co-regulation through the sharing of code and messages among student groups. However, reflections from conducting the course in hybrid mode revealed that collaborative interactions between co-located and online students working on programming problems are more nuanced than in only remote settings, leading to limited use of the intervention. In hybrid settings, interactions span both co-located and remote spaces, as well as synchronous and asynchronous times, creating challenges in maintaining shared awareness and ongoing discourse in collaborative work.
To unpack the findings of the second study, I conducted participatory design sessions with novice students. By directly engaging students in the design, this study aimed to understand their needs for productive collaborative interactions and support in hybrid pedagogical settings for programming. Findings from the sessions revealed the challenges novices face in managing their collaborative work process when programming in hybrid modes. Participants' design solutions offered valuable insights into their preferences for features in programming environments that could effectively support their social regulatory processes in such settings. Based on the findings, a set of design principles have been presented for development of pedagogical tools in hybrid programming learning settings.
Finally, in the third study, the latest intervention was designed to address the need for flexible territories in the programming environment, as identified in the second study. To support novices’ social regulatory behaviour, the intervention encapsulated a collaborative space and a personal space, such that it facilitated osmosis of learning between the two spaces. The evaluation of the intervention in an in-person introductory programming course demonstrated the effect of students' regulatory interactions between these two spaces in supporting students' learning motivation.
The results of my studies show how students' behaviour of socially situating their programming self-regulation leads to the emergence of shared regulatory processes, and how this supports their programming motivation. Based on these findings, I developed design guidelines for collaborative interventions for programming learning classrooms that create supportive conditions for sustaining students' social programming regulation and intrinsic motivation.
This thesis’ results have valuable implications for the design of theory-based interventions to support learning through collaboration, not only in computing education, but also in broader scientific domains.