Scholarlink

This study investigates the efficacy of multimodal AI-generated content (AIGC) tools in design education, focusing on their impact on student behaviors, learning outcomes, and creative performance. Employing a mixed-methods approach, a controlled experiment compared an experimental group (using AIGC-integrated multimodal interfaces) with a control group (traditional pedagogy) across 40 first-year design undergraduates. Quantitative analyses (paired t-tests, ANOVA) and qualitative thematic coding revealed three key findings: (1) Short-term declines in learning motivation (p = 0.001) and self-efficacy (p < 0.001) were observed in the experimental group, attributed to initial tool complexity; (2) Despite motivational challenges, students exhibited high technology acceptance (M = 4.22, p = 0.014) and perceived efficiency gains; (3) AIGC tools mitigated gender disparities in engagement and creativity seen in traditional settings. Qualitative data highlighted tensions between generative outputs’ standardization and creative autonomy. The study proposes actionable optimizations, including phased training modules, customizable generation parameters, and hybrid AI-human workflows. These findings underscore the dual role of AIGC tools as both facilitators of operational efficiency and catalysts for pedagogical innovation, while emphasizing the need for ethical frameworks and longitudinal studies to advance human-AI collaboration in education.

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

Investigating Student Behaviors in Multimodal AIGC-Enhanced Design Education: An HCI-Based Innovative Learning Model

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Jie Ling,
  • Nahua Huang,
  • Zhuohong Ma,
  • Aini Xue,
  • Jiayi Wu,
  • Hongye Li,
  • Zihong Wu,
  • Li Ou-yang

摘要

This study investigates the efficacy of multimodal AI-generated content (AIGC) tools in design education, focusing on their impact on student behaviors, learning outcomes, and creative performance. Employing a mixed-methods approach, a controlled experiment compared an experimental group (using AIGC-integrated multimodal interfaces) with a control group (traditional pedagogy) across 40 first-year design undergraduates. Quantitative analyses (paired t-tests, ANOVA) and qualitative thematic coding revealed three key findings: (1) Short-term declines in learning motivation (p = 0.001) and self-efficacy (p < 0.001) were observed in the experimental group, attributed to initial tool complexity; (2) Despite motivational challenges, students exhibited high technology acceptance (M = 4.22, p = 0.014) and perceived efficiency gains; (3) AIGC tools mitigated gender disparities in engagement and creativity seen in traditional settings. Qualitative data highlighted tensions between generative outputs’ standardization and creative autonomy. The study proposes actionable optimizations, including phased training modules, customizable generation parameters, and hybrid AI-human workflows. These findings underscore the dual role of AIGC tools as both facilitators of operational efficiency and catalysts for pedagogical innovation, while emphasizing the need for ethical frameworks and longitudinal studies to advance human-AI collaboration in education.