Visualization of Physical Education Classroom Interaction Structure and Teaching Improvement Path Based on Social Network Analysis

Existing studies on classroom teaching often focus on unilateral evaluation of teacher behavior or student performance, while insufficient attention is paid to the interaction structure that supports technical skill acquisition. To address this limitation, this study uses social network analysis to visualize and quantify classroom interaction in an 8-hour middle school basketball teaching unit. A total of 1027 interaction events among one teacher and 40 students were collected through non-participant observation and video recording, and a directed weighted adjacency matrix was constructed. UCINET was used to analyze the overall network, cohesive subgroups, and individual centrality. The

Results: show that the interaction network has a low-density (0.28) and high-centrality (0.52) core-periphery structure, with the teacher and several high-skill students occupying central positions. Four homogeneous subgroups based on skill level and gender were identified, and peripheral isolated students showed significantly fewer interaction opportunities. The study proposes teaching improvement paths based on group restructuring and interaction redistribution. The

Method: also provides an engineering-education reference for technical classrooms that use wireless sensing, video analytics, or antenna-supported data capture to monitor learning interaction structures.