The Online with LSE Webinar series recording I watched tonight was one I missed. It was from three years ago and entitled “Applying Graph Theory to Examine the Dynamics of Student Discussions in Small-Group Learning” with Dr. Stanley Lo. The session began with a diagram of the inputs and outputs affecting group work. The study had a few iterations of data collection: taking notes, drawing the sequential order of turns, and combining these two approaches to systematically track who spoke and what was said for mathematical analyses. Lo and the research team used graphs to track talk-turns. Graph theory allows for the quantification of turns. Lo explained that nodes represent people, and centrality is used to measure the activity of someone, for example. The seating arrangement is depicted in the graphs. Interestingly, with a group of eight, the software/analysis was able to identify subgroups. One person seemed to be acting as the bridge between subgroups. They also quantified the number of questions and responses, a normalized talk:ratio based on the “fair” number of turns, and the episode length or sequence of questions and answers. The moderator, Dr. Erin Dolan, asked about the origin of this project. Lo shared how they began with sketches of interactions and this evolved into graphs and trying to quantify the interactions. Lo also explained the difference between graph theory focusing on nodes and connections and how it is different from social network analysis. The math, Lo explained, is very similar, but for graph theory there isn’t the connotation of information flow. The audience asked if Lo had trained students to collect data. Lo is transcribing recordings and from there can analyze the talk-turns. They are now creating a set of scripts in R to automate the process. Lo answered another question that I wondered about: the facilitators set up the chairs, and the patterns affect the dynamics. The methodology took some time to develop, and Lo is now using the data in a pedagogy course for learning assistants. They are now collaborating with a chemistry group using POGIL to quantify interactions! This is an application I would be interested in! Another area that Lo is exploring is studying group dynamics over time. Dolan spoke about how she liked how the article explained three examples with different applications. The group size did seem to affect the appearance of sub groups, with groups with six to eight participants having subgroups. I appreciate Lo’s explanation of the use of example cases in education research to communicate applications. Dolan asked if the methodology could be used to assess whether communication in one group is more equitable. Lo spoke about the difference between participation and attendance and how the qualitative work can help determine what is a contribution. Lo asked: “Can you contribute by not talking?” and provided some examples. I agree. Someone may be actively listening and taking notes…. and may be the notetaker in POGIL. This was an older session and really great to watch!
