Scholarlink

<p>With the rapid advancement of artificial intelligence technologies, generative AI has become increasingly embedded in higher education and demonstrates considerable potential within design pedagogy. Yet, systematic research on the behavioral mechanisms driving design students’ continuous use of generative AI remains limited, especially in relation to influential pathways and nonlinear interaction effects. To address this gap, this study integrates the Unified Theory of Acceptance and Use of Technology (UTAUT), Task–Technology Fit (TTF) theory, and Diffusion of Innovation Theory (DIT) to construct a comprehensive theoretical model. A two-stage analytical strategy combining structural equation modeling (SEM) and artificial neural networks (ANN) is employed to examine linear relationships among variables while simultaneously detecting latent nonlinear influences. Data were collected from 517 design students across multiple Chinese universities. The ANN results identify several variables—such as effort expectancy (EE) and behavioral intention (BI)—that were nonsignificant in the SEM analysis yet exhibited strong predictive importance in the nonlinear model. These findings reveal complex nonlinear mechanisms through which perceptual factors shape students’ willingness to adopt generative AI tools. This study offers two primary contributions. First, it is the inaugural attempt to integrate UTAUT, TTF, and DIT into a unified framework for explaining design students’ continuous intention to use generative AI. Second, it advances methodological practice by implementing a novel two-stage SEM–ANN approach that captures both linear and nonlinear effects, thereby improving the model’s explanatory depth and predictive performance. Overall, the results demonstrate that creative compatibility (CM), relative advantage (RA), observability (OB), and trialability (TR) are especially influential for design students’ sustained adoption of generative AI. The study provides new theoretical and methodological insights for AI-enabled design education and highlights the importance of integrating AI tools in ways that foreground their creative potential and practical value.</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

Exploring Chinese university design students’ continuance intention to use generative AI: a two‑stage SEM–ANN analysis

  • 期刊论文

  • 发表日期：

    2026年1月13日

查看全文
  • Qian Bao,
  • Bin Wang,
  • Na Liang,
  • Jianfeng Wang

摘要

With the rapid advancement of artificial intelligence technologies, generative AI has become increasingly embedded in higher education and demonstrates considerable potential within design pedagogy. Yet, systematic research on the behavioral mechanisms driving design students’ continuous use of generative AI remains limited, especially in relation to influential pathways and nonlinear interaction effects. To address this gap, this study integrates the Unified Theory of Acceptance and Use of Technology (UTAUT), Task–Technology Fit (TTF) theory, and Diffusion of Innovation Theory (DIT) to construct a comprehensive theoretical model. A two-stage analytical strategy combining structural equation modeling (SEM) and artificial neural networks (ANN) is employed to examine linear relationships among variables while simultaneously detecting latent nonlinear influences. Data were collected from 517 design students across multiple Chinese universities. The ANN results identify several variables—such as effort expectancy (EE) and behavioral intention (BI)—that were nonsignificant in the SEM analysis yet exhibited strong predictive importance in the nonlinear model. These findings reveal complex nonlinear mechanisms through which perceptual factors shape students’ willingness to adopt generative AI tools. This study offers two primary contributions. First, it is the inaugural attempt to integrate UTAUT, TTF, and DIT into a unified framework for explaining design students’ continuous intention to use generative AI. Second, it advances methodological practice by implementing a novel two-stage SEM–ANN approach that captures both linear and nonlinear effects, thereby improving the model’s explanatory depth and predictive performance. Overall, the results demonstrate that creative compatibility (CM), relative advantage (RA), observability (OB), and trialability (TR) are especially influential for design students’ sustained adoption of generative AI. The study provides new theoretical and methodological insights for AI-enabled design education and highlights the importance of integrating AI tools in ways that foreground their creative potential and practical value.