Relational Reasoning and Socially Shared Regulation of Learning in Collaborative Problem Solving

Abstract

The ability to solve complex problems in collaborative settings is considered a critical 21st century competency. Yet, national and international reports have revealed deficiencies in both students’ and employees’ teamwork and communication skills, which are essential when working collaboratively. These deficits may be underlain by a limited understanding of how cognitive and social processes operate synchronistically as team members work together to solve complex problems. The current study aimed to investigate how two specific processes—relational reasoning (RR) and socially shared regulation of learning (SSRL)—unfold during a collaborative problem-solving (CPS) task. Specifically, the researcher assessed the extent to which different teams exhibited differential proportions of reasoning and regulation; how team activity was distributed across individuals; and, whether frequent sequences of reasoning and regulation could be identified. To address these aims, four teams of senior undergraduate students (n = 22) were recruited from a capstone design course in mechanical engineering. Over the course of the semester, teams conceptualized and prototyped a design to address a current market need. Each team was video-recorded during the conceptualization process—specifically, as teams evaluated and eliminated ideas from their corpus of designs. Team conversations were transcribed, segmented into utterances, and coded for the presence of RR, SSRL, and task-related and other talk. Results from chi-square tests of independence, social network analysis, and sequence mining revealed that teams indeed exhibited differential proportions of RR and SSRL, with antinomous reasoning and monitoring and control of consensus emerging as key CPS processes. Further, planning and reflection acted as bookends to CPS, while RR and monitoring processes co-occurred in the interim. Finally, CPS alternated between periods of activity that were shared more and less equally among team members. This study contributes to the literature on CPS by exploring the dynamic interplay between RR and SSRL and by demonstrating that CPS can be investigated at the micro level, meso level, and macro level. Methodologically, this study demonstrates how leveraging data mining techniques and assembling compelling visualizations can illustrate the recursive and cyclical character of RR and SSRL. Finally, limitations are noted, and implications for research and practice are forwarded.