Scholarlink

<p>This study aims to investigate the structural characteristics of university students’ artificial intelligence (AI) literacy cognitive networks. Employing Epistemic Network Analysis (ENA), this study modeled the cognitive connections within the thinking processes of 589 pre-service teachers, comparing the network structures formed under an AI course-based instructional context versus a traditional instructional context. The analysis revealed distinct structural features: students in the AI course context developed a more tightly integrated cognitive network, with stronger connections among core dimensions such as knowledge understanding, tool use, and problem-solving. However, ethical judgment and creative ability remained peripheral and under-connected. Further comparative analysis within the AI course context revealed that high-performing students exhibited a more balanced and integrated network across all dimensions, while low-performers’ networks were centered on information screening. Additionally, regression analysis indicated that programming experience was negatively associated with the structural integration of AI literacy within the course context, suggesting a potential mismatch between learners’ prior knowledge and the instructional design. This research offers a structural diagnostic perspective for optimizing AI course design in higher education.</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

Study on the structural characteristics of university students’ AI literacy in AI courses based on epistemic network analysis: the moderating effects of programming experience and gender

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年4月29日

查看全文
  • Zhu Chun,
  • Haibin Chang,
  • Yimin Ning,
  • Yajun Fang

摘要

This study aims to investigate the structural characteristics of university students’ artificial intelligence (AI) literacy cognitive networks. Employing Epistemic Network Analysis (ENA), this study modeled the cognitive connections within the thinking processes of 589 pre-service teachers, comparing the network structures formed under an AI course-based instructional context versus a traditional instructional context. The analysis revealed distinct structural features: students in the AI course context developed a more tightly integrated cognitive network, with stronger connections among core dimensions such as knowledge understanding, tool use, and problem-solving. However, ethical judgment and creative ability remained peripheral and under-connected. Further comparative analysis within the AI course context revealed that high-performing students exhibited a more balanced and integrated network across all dimensions, while low-performers’ networks were centered on information screening. Additionally, regression analysis indicated that programming experience was negatively associated with the structural integration of AI literacy within the course context, suggesting a potential mismatch between learners’ prior knowledge and the instructional design. This research offers a structural diagnostic perspective for optimizing AI course design in higher education.