Using Social Interaction Trace Data to Assess Collaboration Quality and Creative Fluency in Collaborative Design Environments

• In Event: B.E.S.T.: Networks and Social Media

Abstract

Engineering design predominantly takes place in collaborative settings situated in particular contexts. Therefore, it is critical to study both micro-level collaborative interaction processes as well as macro-level spatial and material contexts to promote quality design outcomes and environments. However, research examining interaction dynamics in specific contexts has been limited. Studying interaction dynamics has been difficult due to time and reliability challenges of manual collection and analysis procedures. In this study, we use wearable sociometric badges to automatically and unobtrusively collect conversational and speech dynamics data, including turns taken, successful interrupts, and overlaps. Drawing from situative learning theory, our framework develops two analytic lenses to broaden theoretical insights into engineering and design learning: (a) the social and material context, and (b) the role of activities and interactions. Using a 2x2 design, we randomly assigned engineering graduate students to four conditions varying the material context (Paper vs. Tablet sketching tools) and spatial environment (Commons vs. Standard room). Participants worked collaboratively to generate multiple design ideas for a toy design. Results from 2-stage regression models indicate that while spatial and material contexts did not predict outcomes, social interaction dynamics were highly significant in predicting both collaboration quality and creative fluency. This finding provides important insights about processual factors in collaborative design learning, and for understanding the predictive role of conversation dynamics. We discuss the theoretical contributions to situative learning in collaborative design, the methodological contributions of wearable sensor methods, and the practical contributions to designing collaborative design environments.

Authors

• Ninger Zhou, Purdue U

• Lorraine G. Kisselburgh, Purdue University

• Senthil Chandrasegaran, Purdue U

• Sriram Karthik Badam, Purdue U

• Niklas Elmqvist, Purdue U

• Karthik Ramani, Purdue U