人工智能赋能教育家庭场域背景的公平性研究——基于深圳市C学校23个中小学班级为样本的问卷研究

Research on the Equity of Artificial Intelligence Empowering Education in the Family Sphere—A Questionnaire Study Based on 23 Primary and Middle School Classes from School C in Shenzhen

摘要: 随着人工智能技术的发展，中小学教育得到充分的赋能。不但催生了新的教育与学习方式，也为实现教育公平性带来了新的挑战。为了分析“哪些因素会影响到教育过程公平性”，与“如何影响到教育过程公平性”这两个问题，本文以深圳市C学校23个中小学班级的839名学生及723名家长为样本，探究家庭场域中人工智能赋能教育的过程公平性。研究发现，家长学历与子女AI认知及家长指导行为相关，学历高的家长对AI认知更深、指导倾向性更强，STEM专业背景家长误解率较低；家庭收入影响子女AI使用多样性与家长指导方式，高收入家庭子女AI使用多元性更强。教育过程不公平呈现“认知–行为–情感”三维鸿沟叠加特征，且存在代际传递风险。此外，STEM专业背景家长能缓解学历低带来的认知劣势，而低学历、低收入家庭子女易延续父母在AI认知与使用上的劣势，加剧不公平固化。

Abstract: With the development of artificial intelligence (AI) technology, primary and secondary school education has been fully empowered. This not only gives rise to new educational and learning methods but also brings new challenges to the realization of educational equity. To address the two questions of “which factors affect the equity of the educational process” and “how they affect the equity of the educational process”, this study explores the process equity of AI-empowered education in the family sphere, taking 839 students and 723 parents from 23 primary and secondary school classes of School C in Shenzhen as samples. The findings reveal that parents’ educational background is correlated with their children’s AI cognition and parents’ guidance behavior: parents with higher educational attainment have a deeper understanding of AI and a stronger tendency to provide guidance, while parents with STEM (Science, Technology, Engineering, and Mathematics) backgrounds have a lower rate of misunderstanding AI. Family income influences the diversity of children’s AI usage and parents’ guidance methods, with children from high-income families showing greater diversity in AI use. The inequity in the educational process exhibits the characteristic of an overlapping “cognition-behavior-emotion” three-dimensional gap and carries the risk of intergenerational transmission. Furthermore, parents with STEM backgrounds can mitigate the cognitive disadvantages caused by low educational attainment; however, children from families with low parental education and low income tend to inherit their parents’ disadvantages in AI cognition and usage, which exacerbates the solidification of inequity.