The COVID-19 pandemic has accelerated changes in educational methods, while rapid technological advancements underscore the need for adaptable talent in an evolving society. In this study, we design an experiment and verify the effectiveness of utilizing generative AI tools as a partner based on active learning. In our experiment, we compare the learning experiences and outcomes between participants provided with generative AI tools and those without, within active learning environment. After experiment, participants are asked to conduct a questionnaire to assess motivation and perceived engagement. The results indicated that overall active engagement was higher for those participants who did not use generative AI. However, learning effects were significantly observed when participants engaged deeply with generative AI, such as providing detailed materials, pointing out AI’s mistakes, and sharing prior knowledge. Furthermore, participants utilized generative AI tools as a discussion assistant in the group learning condition, which had a higher rating. In contrast, participants in the individual learning condition had a lower rating, who reported that the interaction with generative AI may disrupt their thinking.

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

Active Learning with Generative AI to Enhance Learning Experience: Experiment Design and Verification

  • Miyuki Ri,
  • Ruichen Cong,
  • Qun Jin

摘要

The COVID-19 pandemic has accelerated changes in educational methods, while rapid technological advancements underscore the need for adaptable talent in an evolving society. In this study, we design an experiment and verify the effectiveness of utilizing generative AI tools as a partner based on active learning. In our experiment, we compare the learning experiences and outcomes between participants provided with generative AI tools and those without, within active learning environment. After experiment, participants are asked to conduct a questionnaire to assess motivation and perceived engagement. The results indicated that overall active engagement was higher for those participants who did not use generative AI. However, learning effects were significantly observed when participants engaged deeply with generative AI, such as providing detailed materials, pointing out AI’s mistakes, and sharing prior knowledge. Furthermore, participants utilized generative AI tools as a discussion assistant in the group learning condition, which had a higher rating. In contrast, participants in the individual learning condition had a lower rating, who reported that the interaction with generative AI may disrupt their thinking.