Teachers’ Behavior in building Agents: Based on Hierarchical Clustering and Thematic Analysis

• In Event: AERA Roundtable Session Sunday 7:45 am - Gold 4
  In Roundtable Session: Behavior and Achievement Across Contexts: Analyzing the Complex Web of Influence Through Mixed Methods

Abstract

This study explored 218 teachers’ behaviors on an intelligent agent creation platform via hierarchical clustering and thematic analysis, identifying three patterns: high browsing-high creation (10.5%), high browsing-low creation (33.5%), and low browsing-low creation (60.0%). Thematic analysis revealed their behavioral logic: the first driven by internal demand or policies; the second following a “browse-reference-apply” chain; the third constrained by environment/time or opting for alternatives. This study provides targeted suggestions for each teacher category.

Authors

• Haiyang Xin, COCOROBO Limited

• Yahui Yu, COCOROBO Limited

• Shuang Li, University of Hong Kong

• Qiannan Niu, COCOROBO Limited

• Lei Gao, The Chinese University of Hong Kong

• Ching Sing Chai, Chinese University of Hong Kong