AI使用方式对大学生批判性思维的影响及群体差异研究

The Impact of AI Usage Patterns on College Students’ Critical Thinking and Group Differences

摘要: 为探究大学生AI工具使用对批判性思维的影响，本研究以全国312名在读本、硕、博学生为对象，通过问卷调查法分析AI工具使用现状、影响因素及批判性思维变化规律。结果显示，大学生AI工具使用呈现高渗透、强依赖、场景集中特征，超六成学生使用ChatGPT、豆包等主流工具，近六成每日使用，核心用于创意写作、作业解答等学习场景，76.28%的学生脱离AI后存在学习障碍，且年级越高、文科/艺术类专业学生依赖度越强。批判性思维变化可通过质疑精神、验证能力、逻辑推理、元认知四大维度评估，其受算法逻辑塑造、认知卸载效应、学科与年级调节三大机制交互影响，呈现初始适应、依赖形成、觉醒与调整三阶段演进特征。研究证实，AI对批判性思维的影响并非绝对，而是取决于使用方式、群体特征与教育引导的综合作用。基于此，本研究从算法赋能、认知重塑、生态重构三维度提出对策，为引导大学生合理使用AI工具、培养批判性思维提供参考。

Abstract: To explore the impact of college students’ use of AI tools on their critical thinking, this study surveyed 312 undergraduate, master’s, and doctoral students nationwide. Through a questionnaire-based approach, it analyzed the current status of AI tool usage, influencing factors, and the changing patterns of critical thinking. The results indicate that AI tool usage among college students is characterized by high penetration, strong dependence, and concentrated application scenarios. Over 60% of students use mainstream tools such as ChatGPT and Doubao, nearly 60% use them daily, primarily for learning scenarios including creative writing and homework problem-solving. Additionally, 76.28% of students experience learning barriers when disconnected from AI, with dependence being stronger among higher-grade students and those majoring in liberal arts or art. Changes in critical thinking can be evaluated through four dimensions: questioning spirit, verification ability, logical reasoning, and metacognition. These changes are interactively influenced by three mechanisms—algorithm logic shaping, cognitive offloading effect, and regulation by academic discipline and grade—and exhibit an evolutionary trajectory comprising three stages: initial adaptation, dependence formation, and awakening and adjustment. The study confirms that the impact of AI on critical thinking is not absolute but depends on the combined effects of usage patterns, group characteristics, and educational guidance. Based on these findings, this research proposes countermeasures from three dimensions: algorithm empowerment, cognitive restructuring, and ecological reconstruction, aiming to provide references for guiding college students to use AI tools rationally and fostering their critical thinking.