Unpacking the talk: investigating how multiparty face-to-face collaborative discourse forges learning and collaboration

• Main Author: Chua, Victoria Yi Han
• Other Authors: Suzy Styles
• Format: Thesis-Master by Research
• Language: English
• Published: Nanyang Technological University 2024
• Subjects: Social Sciences; Collaborative learning; Collaborative discourse; Speech analysis
• Online Access: https://hdl.handle.net/10356/178034
• Institution: Nanyang Technological University

As the use of collaborative practices in work and education becomes common, it has become imperative to understand collaborative experiences and its underlying processes. Accordingly, many studies have attempted to delineate behaviours of successful collaboration, and methods to extract these behaviours from collaborative discourse. However, gaps remain in understanding how these behaviours emerge and can be detected in co-located multi-party collaboration. In Chapter 1, we identified participation equity, reciprocal interaction, mutual understanding, information density and emotional rapport as important constructs of productive collaboration. We highlighted limitations in attempting to leverage audio-visual features extracted from collaborative discourse for monitoring of collaboration. Subsequently, in Chapter 2, we adapted dyadic measures of the five constructs to a multi-party co-located collaborative learning context. To understand the unique effects of individual’s contributions and their team members’ contributions respectively, we defined comparable individual-level and group-level metrics of each construct. We then provided a descriptive account of how students varied in the ways they coordinated their speech, vocabulary and laughter when working in an authentic classroom setting. In Chapter 3, we systematically investigated if participation equity (Study 1), reciprocal interaction (Study 2), mutual understanding (Study 3), information density (Study 4) and emotional rapport (Study 5), impact co-located multi-party collaborative outcomes. Perceived collaboration quality only varied on the function of reciprocal interaction at the group level. Exploratory analyses revealed that the knowledge level of the highest performing individual affected individual learning gains of the remaining team members. Thus, the role of high performing individuals should be considered more thoroughly in evaluating collaborative dynamics and outcomes. Finally in Chapter 4, we summarise how the current work has implications for researchers interested in collaboration within educational and work domains as well as educators implementing collaborative learning in their classrooms.