大学生人工智能使用行为与焦虑水平的关联研究
The Correlation Between College Students' Artificial Intelligence Usage Behavior and Anxiety Level
提交：2026-03-05录用：2026-04-13出版：2026-05-11
环球医学进展, 2026, 5(1), 20-25; https://doi.org/10.58244/glmedp.253737
摘要：目的：探究大学生人工智能（AI）使用行为与焦虑水平的多维度关联，分析生成式AI与决策式AI的差异化影响，揭示“技术-个体-环境”三者的交互机制，为大学生心理健康干预与AI智慧教育应用提供科学依据。方法：采用方便抽样法，选取杭州师范大学全日制本科生为研究对象，通过问卷星平台发放《AI使用行为多维因素调查问卷》、基本信息问卷及GAD-7焦虑量表，运用SPSS 27.0进【更多...】
关键词：大学生；人工智能使用行为；焦虑水平；中介效应；生成式AI；决策式AI
关键词： 大学生；人工智能使用行为；焦虑水平；中介效应；生成式AI；决策式AI
摘要
摘 要：
目的：探究大学生人工智能（AI）使用行为与焦虑水平的多维度关联，分析生成式AI与决策式AI的差异化影响，揭示“技术-个体-环境”三者的交互机制，为大学生心理健康干预与AI智慧教育应用提供科学依据。方法：采用方便抽样法，选取杭州师范大学全日制本科生为研究对象，通过问卷星平台发放《AI使用行为多维因素调查问卷》、基本信息问卷及GAD-7焦虑量表，运用SPSS 27.0进行信效度检验、相关分析与回归分析。结果：最终回收有效问卷168份，验证《AI使用行为多维因素调查问卷》具有良好信效度（Cronbach’s α系数＞0.8）；大学生AI日均使用时长、依赖程度与焦虑水平呈正相关（P＜0.05）；生成式AI的过度依赖对焦虑的预测作用显著（β=0.32，P＜0.01），决策式AI的使用不安感与焦虑水平呈正相关（r=0.28，P＜0.01）；技术适应性、自我效能感在AI使用行为与焦虑间起部分中介作用。结论 大学生AI使用行为与焦虑水平存在密切关联，不同类型AI的影响路径存在差异，需通过规范AI使用引导、提升技术适应能力等方式促进大学生数字环境下的心理适应。
目的：探究大学生人工智能（AI）使用行为与焦虑水平的多维度关联，分析生成式AI与决策式AI的差异化影响，揭示“技术-个体-环境”三者的交互机制，为大学生心理健康干预与AI智慧教育应用提供科学依据。方法：采用方便抽样法，选取杭州师范大学全日制本科生为研究对象，通过问卷星平台发放《AI使用行为多维因素调查问卷》、基本信息问卷及GAD-7焦虑量表，运用SPSS 27.0进行信效度检验、相关分析与回归分析。结果：最终回收有效问卷168份，验证《AI使用行为多维因素调查问卷》具有良好信效度（Cronbach’s α系数＞0.8）；大学生AI日均使用时长、依赖程度与焦虑水平呈正相关（P＜0.05）；生成式AI的过度依赖对焦虑的预测作用显著（β=0.32，P＜0.01），决策式AI的使用不安感与焦虑水平呈正相关（r=0.28，P＜0.01）；技术适应性、自我效能感在AI使用行为与焦虑间起部分中介作用。结论 大学生AI使用行为与焦虑水平存在密切关联，不同类型AI的影响路径存在差异，需通过规范AI使用引导、提升技术适应能力等方式促进大学生数字环境下的心理适应。
关键词：大学生；人工智能使用行为；焦虑水平；中介效应；生成式AI；决策式AI
Abstract:
Objective: To explore the multi-dimensional correlation between college students' artificial intelligence (AI) usage behavior and anxiety level, analyze the differential impacts of generative AI and decision-making AI, reveal the interaction mechanism among "technology-individual-environment", and provide a scientific basis for college students' mental health intervention and the application of AI intelligent education. Methods: Convenience sampling was adopted to select full-time undergraduates from Hangzhou Normal University as the research objects. Questionnaires including the Multidimensional Factor Questionnaire of AI Usage Behavior, the Basic Information Questionnaire and the GAD-7 Anxiety Scale were distributed through the Wenjuanxing platform. SPSS 27.0 was used for reliability and validity test, correlation analysis and regression analysis. Results: A total of 168 valid questionnaires were recovered, verifying that the Multidimensional Factor Questionnaire of AI Usage Behavior had good reliability and validity (Cronbach’s α coefficient > 0.8); college students' daily AI usage duration and dependence were positively correlated with anxiety level (P < 0.05); excessive dependence on generative AI had a significant predictive effect on anxiety (β=0.32, P < 0.01); the sense of unease in using decision-making AI was positively correlated with anxiety level (r=0.28, P < 0.01); technical adaptability and self-efficacy played a partial mediating role between AI usage behavior and anxiety. Conclusion: There is a close correlation between college students' AI usage behavior and anxiety level, and the impact paths of different types of AI are different. It is necessary to promote college students' psychological adaptation in the digital environment by standardizing AI usage guidance and improving technical adaptability.
Objective: To explore the multi-dimensional correlation between college students' artificial intelligence (AI) usage behavior and anxiety level, analyze the differential impacts of generative AI and decision-making AI, reveal the interaction mechanism among "technology-individual-environment", and provide a scientific basis for college students' mental health intervention and the application of AI intelligent education. Methods: Convenience sampling was adopted to select full-time undergraduates from Hangzhou Normal University as the research objects. Questionnaires including the Multidimensional Factor Questionnaire of AI Usage Behavior, the Basic Information Questionnaire and the GAD-7 Anxiety Scale were distributed through the Wenjuanxing platform. SPSS 27.0 was used for reliability and validity test, correlation analysis and regression analysis. Results: A total of 168 valid questionnaires were recovered, verifying that the Multidimensional Factor Questionnaire of AI Usage Behavior had good reliability and validity (Cronbach’s α coefficient > 0.8); college students' daily AI usage duration and dependence were positively correlated with anxiety level (P < 0.05); excessive dependence on generative AI had a significant predictive effect on anxiety (β=0.32, P < 0.01); the sense of unease in using decision-making AI was positively correlated with anxiety level (r=0.28, P < 0.01); technical adaptability and self-efficacy played a partial mediating role between AI usage behavior and anxiety. Conclusion: There is a close correlation between college students' AI usage behavior and anxiety level, and the impact paths of different types of AI are different. It is necessary to promote college students' psychological adaptation in the digital environment by standardizing AI usage guidance and improving technical adaptability.
Keywords: College students; Artificial intelligence usage behavior; Anxiety level; Mediating effect; Generative AI; Decision-making AI
--
正文
此为预上线版本，正文内容请期待后续更新。
参考文献
参考文献 / References：
- 张正, 李明, 王芳. 生成式人工智能对大学生认知能力的影响研究[J]. 电化教育研究, 2024, 45(3): 89-95.
- Li J, Wang Y, Zhang H. A multidimensional model of AI anxiety based on fear acquisition theory[J]. Journal of Psychosocial Research on Cyberspace, 2023, 8(2): 45-62.
- Fitzpatrick K, Smith M, Jones R. Algorithmic bias, information overload, and anxiety among college students: A cross-sectional study[J]. Computers in Human Behavior, 2022, 131: 107189.
- Lupton D. AI dependence and anxiety: Exploring the reciprocal relationship[J]. Social Science & Medicine, 2021, 285: 113768.
- Agca H, Yilmaz S, Karaman K. The dual effects of AI integration in education: Reducing learning anxiety while increasing career anxiety[J]. Journal of Educational Technology & Society, 2020, 23(4): 210-223.
- Heinssen R K, Glass C R, Knight L A. The computer anxiety rating scale: Development, reliability, and validity[J]. Journal of Counseling Psychology, 1987, 34(4): 387-394.
- Brosnan M J. The computer anxiety rating scale: A replication and extension[J]. Journal of Educational Computing Research, 1998, 18(2): 169-180.
- 中华人民共和国国家互联网信息办公室. 新一代人工智能治理原则[Z]. 2019.
- 中国心理学会. 2022年国民心理健康蓝皮书[R]. 北京: 科学出版社, 2023.
