Productive interactions in engineering practice: A comparison between student and expert teams using network analysis

This study is prepared for presentation only (no full paper). Engineering capstone projects provide students the opportunity to prepare for their careers by providing a messy, open-ended context where they do real engineering work. These projects are commonly seen as a venue for students to integrate and apply the foundational knowledge and skills that they have learned in their coursework. In this study, we examine interactions of teams completing a capstone project through the lens of disciplinary practice. We use process data to compare the interactions of three student teams and an expert team as they develop a process to deposit silicon nitride thin films using a virtual chemical vapor deposition (CVD) reactor. Borrowing from Pickering's (1995) work in science education, we categorize engineering practice according to conceptual, material, and social aspects. We then apply Damsa's (2014) construct of productive interactions to analyze the teams' discourse. Damsa's two dimensions of productive interactions, epistemic and regulative, align with the conceptual aspect of practice and the social aspect of practice, respectively. To capture the material aspect of practice, we have added a third type of productive interactions, experimental. This study examines in depth three teams of undergraduate engineering students and a team of expert engineers. This project has the participants take on the role of a team of process engineers who are tasked with optimizing a virtual CVD reactor. The simulation provides data representative of an industrially sized reactor, including common and special causes of variation. As in industry, each experiment and measurement costs the team money. During this three-week project, teams had weekly meetings with a faculty member who acted in the role of the project supervisor. The first meeting served as an authorization meeting where teams met with the supervisor to have their initial experimental parameters, experimental strategy, and budget approved before they were granted access to the reactor. Teams were audio recorded whenever two or more members met during the project, and all audio was transcribed. Two meeting segments of approximately one-hour each were analyzed for each team (eight total) - one from just before the authorization meeting and another from between each team's first and second experiments. Open coding led to seventeen productive interactions codes, which aligned with the three aspects of engineering practice. We then used Epistemic Network Analysis (ENA) to identify relations between participant interactions and aspects of practice. The findings reveal that interactions related to the social aspect of practice have the highest number of occurrence for all teams. This result emphasizes the importance of technical coordination in engineering practice. Results from ENA reveal consistently strong ties between the conceptual and material aspect of practice for the expert team while the student teams show episodic but not consistent ties between material and conceptual aspects. It appears that the experts use engineering tools to mediate their sense-making processes in a way that directs their experimental strategy. Students tended to become more immersed in first-principles, often in ways that did not directly apply to the experiments they were designing and analyzing.