Scholarlink

<p>Video lectures are widely used learning tools, providing flexibility beyond the constraints of time and space. AI chatbots have emerged as a promising solution to facilitate students’ social interactions in video lectures. This study conducted an eye-tracking experiment with 99 university students (88 female, 11 male) randomly assigned to three groups to examine how the design of AI chatbots’ conversational paths (a user-initiated path vs. a mind-wandering-related proactive path vs. a content-related proactive path) affect students’ social and cognitive outcomes. The findings highlight the advantages of proactive chatbot paths over user-initiated designs, showing improvements in social presence (η² = 0.17), motivation (η² = 0.10), and retention (η² = 0.19). This study contributes to the literature by distinguishing the effects of different proactive conversational paths and clarifying the mechanisms through which proactive AI interactions support learning. The results also provide practical implications for designing video lectures with AI chatbots: developers and educators are encouraged to align proactive AI interactions with learning materials, instructional goals, and learner characteristics, rather than assuming universal benefits.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Enhancing learning from video lectures with proactive AI chatbots

  • 期刊论文

  • 发表日期：

    2026年7月27日

查看全文
  • Zhongling Pi,
  • Dan Zhou,
  • Yuxuan Wei,
  • Siyu Zheng,
  • Xiying Li

摘要

Video lectures are widely used learning tools, providing flexibility beyond the constraints of time and space. AI chatbots have emerged as a promising solution to facilitate students’ social interactions in video lectures. This study conducted an eye-tracking experiment with 99 university students (88 female, 11 male) randomly assigned to three groups to examine how the design of AI chatbots’ conversational paths (a user-initiated path vs. a mind-wandering-related proactive path vs. a content-related proactive path) affect students’ social and cognitive outcomes. The findings highlight the advantages of proactive chatbot paths over user-initiated designs, showing improvements in social presence (η² = 0.17), motivation (η² = 0.10), and retention (η² = 0.19). This study contributes to the literature by distinguishing the effects of different proactive conversational paths and clarifying the mechanisms through which proactive AI interactions support learning. The results also provide practical implications for designing video lectures with AI chatbots: developers and educators are encouraged to align proactive AI interactions with learning materials, instructional goals, and learner characteristics, rather than assuming universal benefits.