Learner Help-Seeking Behaviors Toward Intelligent Agents in MOOC Settings

Help-Seeking is a crucial behavior in the learning process, particularly within online learning environments, where the strategies learners employ to seek assistance can significantly impact their learning outcomes. This paper explores help-seeking behaviors in MOOCs, with a specific focusing on interactions with the intelligent agent "Xiaomu" in a Data Structures course from XuetangX. By analyzing interaction data from 4,751 learners, we identified six distinct types of questions posed to the intelligent agent, which revealed notable gender differences in usage frequency. Female learners exhibited higher interaction rates, and learners enrolled in a greater number of courses tended to pose a broader range of questions. This study illuminates learner behaviors in MOOC settings, providing empirical evidence to inform the design and functionality of intelligent agents.