北京高校学生AI大模型持续使用意愿影响因素研究

Research on the Influencing Factors of Continuance Usage Intention of AI Large Models among Beijing College Students

摘要: 本文以北京高校学生为研究对象，旨在探究其AI大模型的使用现状及持续使用意愿的影响因素，为提升学生AI素养和优化教育场景下AI应用提供实证支撑。研究基于技术接受模型(TAM)与期望确认模型(ECM)，将影响因素区分为内部因素与外部因素，并采用描述性统计、信效度检验、相关性分析和多元线性回归对数据进行分析。结果显示，ChatGPT与豆包为最常用模型，写作辅助与语言翻译为主要应用场景，回答不精准和复杂问题处理不足为突出问题；回归模型显著，内部因素(β = 0.614)与外部因素(β = 0.374)均正向影响持续使用意愿，且内部因素作用更强。研究表明，北京高校学生AI使用已较普及但应用层次偏浅，提升使用意愿需重点优化产品性能，尤以准确性与响应质量为优先。研究可为深化高校AI应用及提升学生数字素养提供实证参考。

Abstract: This study takes Beijing college students as the research subjects, aiming to investigate their current usage of AI large models and the factors influencing their continuance usage intention, so as to provide empirical support for enhancing students’ AI literacy and optimizing AI applications in educational contexts. Grounded in the Technology Acceptance Model (TAM) and the Expectation-Confirmation Model (ECM), the study categorizes influencing factors into internal and external factors, and employs descriptive statistics, reliability and validity testing, correlation analysis, and multiple linear regression for data analysis. The results show that ChatGPT and Doubao are the most frequently used models, with writing assistance and language translation being the primary application scenarios, while inaccurate responses and inadequate handling of complex problems are prominent issues. The regression model is significant, with both internal factors (β = 0.614) and external factors (β = 0.374) positively affecting continuance usage intention, and internal factors exerting a stronger influence. The findings indicate that AI usage among Beijing college students is relatively widespread but remains shallow in terms of application depth; enhancing usage intention requires prioritizing product performance optimization, particularly in terms of accuracy and response quality. This study provides empirical references for deepening AI applications in higher education and improving students’ digital literacy.