Using Social Network Analysis to Examine Interactional Patterns in Scientific Argumentation

• In Event: Advancing Students' Argumentation: Moving Beyond Structure

Abstract

To transform discussions in science classrooms towards including argumentation, we need a stronger understanding of what argumentative discourse patterns look like, paying particular attention to teacher and student roles (Kuhn & Reiser, 2006). While the structural elements of an argument have been studied in many ways (e.g. Authors, 2006; Sandoval & Millwood, 2005), these analyses do not capture the highly interactive aspects of this practice, including critiquing and debating the strength of a claim with others, as well as evaluating and revising claims (Ford, 2008). As such, research is needed to understand how to support students in critiquing one another, as well as building off of the ideas of their peers as they engage in the process of argumentation (Andriessen, 2007). Thus, our study is guided by the question: How can social network analysis (SNA) be used to examine interactional patterns in argumentative talk?
We use SNA to examine the interaction patterns within a science seminar activity of three classrooms enacting a pilot curricula (Regents of the University of California, 2012a; 2012b; 2013) focused on argumentation. A science seminar is a whole class activity in which students orally debate explanations to a question. SNA is a methodology that seeks to identify underlying patterns of social relations based on the way actors in a network are related to one another (Scott, 1991; Wasserman & Faust, 1997). A classroom can be construed as a social network in which the teacher and students are actors (Borgatti & Ofem, 2010). Using science seminar transcripts, we constructed sociograms, visual representations of social networks (see Figure 1), which provide insight into the interactional patterns during the science seminars.
These sociograms show whether interactions are occurring between all actors or whether some actors are communicating more, or less, with group members (Haythornthwaite, 2002). Sociograms consist of nodes, and ties that connect the nodes. The nodes are the dots in the figure, which in our study represent a teacher or a student, while the ties capture the interaction being focused on (e.g. asking a question). From the sociograms, we identify participants who are being positioned in interesting ways in the network. For example, this includes people at the periphery of the network, central actors, or even people serving as bridges between participants and the rest of the group (see circled nodes in Figure 1). Further, sociograms shed light on the “flow” of information that is exchanged between actors (Thorpe, McCormick, Kubiak & Carmichael, 2007). Thus, SNA allows us to study: the nature of the interaction patterns, the roles participants take on, and the ways classroom members share, construct and revise knowledge as they engage in argumentative discourse. These interaction patters and roles can help us to identify and develop instructional strategies that facilitate shifts in discourse norms (Kuhn & Reiser, 2006), such as assigning students to explicit roles like “critiquer” and “synthesizer” that increase their centrality in the discourse.

Authors

• Maria Gonzalez-Howard, Boston College

• Katherine L. McNeill, Boston College