Network Formation and Interaction Patterns in a Statistics Classroom

Presented by:
Siming Su & Yuning Li (University of California, Santa Barbara)

Social interactions in classrooms are a major part of students' college life. Literature shows that recognition from peers forms an important part of a student's identity. Research has also shown gender biases exist in students’ recognition of strong(in course material) peers. We analyze survey data collected from students(n = 287) in an undergraduate classroom at two points in the term. We use Social Network Analysis(SNA) and ERGM’s (Exponential Random Graph Model) to investigate students’ interactions, attitudes, and perceptions of strong peers. We discuss our results which include significant relationships between students being perceived as strong by their peers and student grades (midterm: p-value= 0.0167; final: p-value= 0.0201), homophily of the interaction network with respect to certain grades, certain races, and underrepresented minority groups* (URM’s). Using ERGM we see that students who are perceived as “strong students''(knowledgeable in course material) by their peers interact more with others. In conclusion, using SNA and ERGM we have analyzed and provided insights into the network of interactions and perceptions that forms in a statistics classroom.