Deeper Learning Through Conflict Resolution or Stubborn Resistance to Change?

• In Event: 52.027 - Fostering Deep Learning in Problem-Solving Contexts Through Effective Design of Learning Environments With Technology Support

Abstract

Objectives
Healthcare professionals rely heavily on teams to achieve effective patient care and safety; therefore, engaging students in collaborative problem solving constitutes an essential component of health professions education (Weaver, Dy, & Rosen, 2014). This paper extends the investigation of deep learning beyond learning as an individual endeavor by looking at the collaborative achievement of teams. Although learners might indicate an agreement to a collective solution, deeper learning occurs only after they integrate group reasoning into their individual conceptions (Jeong & Chi, 2007). These changes in knowledge representation can be evaluated through the measure of knowledge convergence (KC), a similarity in team members’ post-collaboration knowledge representations derived through interaction (Fischer & Mandl, 2005). While previous studies have statistically associated degree of KC with achievement of learning outcomes, there still exists a significant lack of research investigating factors influencing its attainment.

This study focused specifically on the impact of learners’ conflict resolution on their deep learning as measured by KC. Conflict-oriented consensus building was listed as one of the major categories in Weinberger and Fischer’s (2006) framework analyzing knowledge construction discourse. It indicates learners’ disagreement with or repairs to peers’ contributions. Weinberger and Fischer argued that, in comparison to the remaining categories, conflict-oriented consensus building was most supportive of learners’ active engagement with and subsequent appropriation of peers’ ideas. However, while prior research (e.g. Fischer, Bruhn, Grasel, & Mandl, 2002) hints at possible positive impact from conflict-oriented consensus building on learners’ KC, no statistical analysis has been conducted to directly investigate their relationship.

In our study, 15 American undergraduates enrolled in a radiographic physics course engaged in collaborative problem solving. In teams of three, they used CmapTools software to draw a concept map identifying possible causes of an x-ray machine breakdown. Learners also submitted an individual map to solve the same problem prior to their collaboration and revised it afterwards. Data sources for this mixed-method study included learners’ maps and also recordings of collaborative sessions and post-collaboration interviews with individual students, who participated on a voluntary basis.

Pathfinder analysis was conducted to compare a student’s post-collaboration individual map with the other team member’s revised maps. The similarity scores yielded from these two comparisons were averaged to obtain the student’s KC score. Triads’ discourses were transcribed and parsed into 2,211 conversational turns, which were coded by two researchers (Cohen’s Kappa = .78) using the above-mentioned Weinberger and Fischer’s (2006) framework.

Regression analysis with errors clustered under triads was conducted. Contrary to prior research findings, increased conflict negotiation behaviors by an individual resulted in fewer similarities between his or her second map and other members’ maps (ρ= .013). Qualitative analysis of learners’ collaboration process and interview transcripts was also conducted, revealing interaction patterns and strategies employed during learners’ conflict negotiation that had impacted their attainment of KC. These findings highlighted the roles of self-confidence, openness to change, and team dynamics to the attainment of KC. Future research regarding instructional support of these affective factors is needed.

Authors

• Weichao Chen, University of Virginia

• Carla M. Allen, University of Missouri - Columbia

• David H. Jonassen, University of Missouri - Columbia