Characterizing social behavior patterns in teaching assistant interactions with students

Understanding how relationships between instructors and students develop is important for understanding the undergraduate student experience. We expect the development of positive relationships is related to the social practices (e.g., greetings, using names, sympathizing, or empathizing with students) that instructors use in the course of normal classroom interactions with students. We recorded interactions between instructors and students in remote synchronous online physics problem-solving sessions and surveyed students about their perceptions of their instructors. We selected the highest-rated instructor and lowest-rated instructor in our sample and identified social practices in their conversations with students. We first characterized the frequency of social practice usage by each instructor in their conversations with students. We find that both instructors relied on a set of core social practices in most conversations with students, but that our higher-rated instructor used comparatively more positive commentary and sympathizing or empathizing behaviors than our lower-rated instructor. In comparison, our lower-rated instructor engaged in more negative commentary. Using network analysis, we then explored patterns in co-occurrences of social practices used by each instructor moment-to-moment in conversations and compared the instructors’ social practice network patterns. We find that our higher rated-instructor used a greater variety of social practices during moment-to-moment interactions with students, while our lower-rated instructor spent most of his time focused on classroom business. We suggest that professional development for instructors should include guidance on how messages are delivered in classes and encourage the use of high-impact social practices to foster positive relationships with students.

Figure 1

Schematic of stanza structure and code co-occurrence vector analysis.

Figure 2

Network graphs characterizing co-occurrence of social practice use for H (top left, red) and L (top right, green). Line weight represents the frequency with which two practices appeared in the same stanza. If viewing these plots electronically, we recommend zooming in on the plots to better see relative line weights. The reference network graph consisting of two nodes and a single line (bottom right, black) represents two hypothetical codes if they appeared together within every stanza. The bottom left graph was generated by subtracting L’s graph from H’s graph. In this graph, line weight represents the magnitude of the difference in line weights and the color represents which instructor’s graph had a higher weight.