University Students' Continued Use of Survey Data for Generative AI.xlsx

The growing adoption of generative AI (GenAI) in education underscores the importance of identifying the factors that shape university students' intentions to sustain their use of this technology. Despite its growing adoption, limited research has investigated how the attributes of GenAI shape students' sustained usage intentions. Grounded in the Heuristic-Systematic Model and Means-End-Chain Theory, this study conducted an online survey of university students in mainland China through random sampling (time1, n = 823; time2, n = 712), explores how heuristic factors (anthropomorphism and likability) and systematic factors (perceived intelligence and AI compatibility) influence students' intention to continue using GenAI, with utilitarian and hedonic values serving as mediators. The findings indicate that anthropomorphism, likability, perceived intelligence, and AI compatibility have a positive impact on students' intention to continue using GenAI. Moreover, utilitarian and hedonic values serve as mediators, linking these attributes to behavioral intentions, highlighting the dual role of value perceptions in driving continued use. The study highlights the fear of AI as a negative moderating factor that weakens the direct effects of these attributes on intention, emphasizing the critical role of individual characteristics in shaping human-AI interactions. This study enhances the literature on human-AI interaction by combining heuristic and systematic cues with user value perceptions to clarify the mechanisms driving the adoption and continued use of GenAI. The findings offer theoretical and practical insights, emphasizing the need for thoughtful design of GenAI attributes to enhance its utility and user experience while addressing concerns related to fear of AI to promote long-term engagement.生成式AI (Generative AI)在教育领域的应用日益普及，凸显出识别影响大学生持续使用该技术意愿的各类因素的重要性。尽管GenAI的应用普及率持续提升，但目前针对其自身属性如何影响学生持续使用意愿的研究仍较为匮乏。本研究以启发式-系统式模型 (Heuristic-Systematic Model)与手段-目的链理论 (Means-End-Chain Theory)为理论基础，通过随机抽样方式对中国大陆地区大学生开展两轮线上问卷调查（第一轮：n=823；第二轮：n=712），探究启发式因素（拟人化 (anthropomorphism)与好感度 (likability)）与系统式因素（感知智能 (perceived intelligence)与AI兼容性 (AI compatibility)）对大学生持续使用GenAI意愿的影响路径，并以功利价值与享乐价值作为中介变量。研究结果显示，拟人化、好感度、感知智能与AI兼容性均对大学生持续使用GenAI的意愿产生显著正向影响。此外，功利价值与享乐价值作为中介变量，将上述GenAI属性与用户行为意愿建立关联，凸显了价值感知在推动持续使用行为中的双重作用。本研究同时指出，AI恐惧心理作为负向调节变量，会削弱上述GenAI属性对使用意愿的直接影响，凸显了个体特征在人机交互过程中的关键作用。本研究将启发式与系统式线索与用户价值感知相结合，厘清了驱动GenAI采纳与持续使用的内在机制，丰富了人机交互领域的相关研究文献。本研究结果兼具理论与实践价值，提出应通过精细化设计GenAI的相关属性，以提升其使用效用与用户体验，同时针对性缓解用户的AI恐惧心理，从而促进用户的长期持续使用。