【作 者】：
马银琦，黄 恒，毋 磊，黄 丹
【关 键 词 】：
生成式人工智能； 研究生； 科研创新力； 技术赋能； 工具依赖
【栏 目】：
学习环境与资源
【中文摘要】：
生成式人工智能（AIGC）在教育领域展现出巨大潜力，也为研究生科研创新力培养带来新的机遇和挑战。基于1714份研究生调查数据，采用普通最小二乘法、负二项回归、熵平衡匹配法及无条件分位数回归等方法，探讨AIGC对研究生科研创新力的影响。研究发现：近九成研究生使用过AIGC辅助科研，主要将其用于语言润色和文献查阅；使用AIGC显著提升了研究生的科研创新力，尤其是在科研创新思维、科研实践能力和科研创新品质方面；AIGC使用频率对研究生科研创新力有显著的促进作用，且与研究生科研创新成果之间存在显著的非线性关系；使用AIGC及其使用频率对研究生科研创新力的影响在不同群体间的异质性明显。据此，应增强信息技术指导，提高研究生使用AIGC的广度与深度，倡导合理使用AIGC并遵守伦理规范，缩小数字鸿沟与优化个性化学习。
【英文摘要】：
Generative Artificial Intelligence (GAI) has demonstrated great potential in the field of education and brought new opportunities and challenges for the cultivation of graduate students' research and innovation abilities. Based on the survey data from 1,714 graduate students, the impact of GAI on graduate students' research and innovation abilities was investigated through methods such as Ordinary Least Squares, Negative Binomial Regression, Entropy Balancing Matching, and Unconditional Quantile Regression. It is found that nearly 90% of graduate students have used GAI to assist their research, primarily for language polishing and literature review. The use of GAI has significantly enhanced graduate students' research and innovation abilities, particularly in innovative thinking, practice ability, and innovation quality. The frequency of GAI use significantly contributes to graduate students' research and innovation abilities, and has a significant nonlinear relationship with graduate students' innovative research outcomes. The impact of using GAI and its frequency on graduate students' research and innovation abilities is significantly heterogeneous among different groups. Therefore, it is necessary to strengthen the guidance of information technology, improve the breadth and depth of graduate students' use of GAI, advocate for advocate the rational use of GAI and adherence to ethical norms, narrow the digital divide and optimize personalized learning.