华东师范大学学报(教育科学版) ›› 2026, Vol. 44 ›› Issue (7): 61-75.doi: 10.16382/j.cnki.1000-5560.2026.07.005

• 基础教育 • 上一篇    下一篇

是什么塑造了中小学生的创新能力——基于机器学习方法的实证研究

钱佳, 崔晓楠   

  1. 华中师范大学教育学院,武汉 430079
  • 出版日期:2026-07-01 发布日期:2026-06-23
  • 基金资助:
    中国教育学会“师范教育协同提质计划”专项重大课题“乡村教育振兴视野下师范教育跨界治理效能与过程评价研究”(202400002204ZXA)。

What Shapes the Innovative Ability of Primary and Secondary School Students: An Empirical Study Based on Machine Learning

Jia Qian, Xiaonan Cui   

  1. School of Education, Central China Normal University, Wuhan 430079, China
  • Online:2026-07-01 Published:2026-06-23

摘要:

创新人才早期培养的关键在于以适宜的教育促进中小学生创新能力发展。为发掘并培养拔尖创新后备人才、夯实人才培养的“底座”,基于7552名中小学生调查数据,聚焦创新能力内外部成分异质性视角,以机器学习方法实证考察学生创新能力的发展机制。研究发现:第一,学生创新能力培养是各类内外部因素长期交互影响的复杂过程,在此过程中,个体内部成分的贡献率(83.09%)大于外部环境成分(13.44%)。第二,个体内部成分中的学习兴趣和乐观特质等因素最为关键,而外部环境成分中教师支持的贡献率最高。第三,从创新能力的发展阶段来看,各因素的贡献率存在时序波动性,其中学生韧性特质和闲暇时间随着学段的延伸愈发关键,教师支持在小学和初中学段更为重要。第四,从创新能力的预测模式来看,知识基础、内在动机、人际支持、闲暇时间均与创新能力之间呈现非线性关系。对此,应坚持“全纳+分层”相结合的学生创新能力培养理念,完善“内生+外驱”双赋能的培养机制,构建“分段+贯通”互衔接的培养模式,让拔尖创新人才不断涌现。

关键词: 学生创新能力, 拔尖创新人才早期培养, 机器学习方法, 影响因素

Abstract:

The key to the early cultivation of innovative talents is to promote the development of primary and secondary school students’ innovative ability through appropriate education. In order to discover and cultivate the top innovative talents and consolidate the foundation of talent cultivation, based on the survey data of 7552 students, the study focuses on the heterogeneity of internal and external components of the innovative ability, and empirically examines the development mechanism of the students’ innovative ability by machine learning. The study found that, first, the development of students’ innovative ability is a complex process of long-term interaction between various internal and external factors, in which the contribution rate of the internal component of an individual (83.09%) is greater than that of the external environment component (13.44%). Second, the individual’s internal component, such as learning interest and optimism, is the most critical, while the external environment component, teacher support, has the highest contribution rate. Third, in terms of the stage of development of innovativeness, there is temporal fluctuation in the contribution of the factors. Among them, students’ resilience trait and leisure time become more and more critical with the extension of school stages, and teacher support is more important in primary and junior high school. Finally, from the perspective of the prediction model of innovation ability, there is a non-linear relationship between knowledge base, intrinsic motivation, interpersonal support and leisure time and innovation ability. In this regard, we should adhere to the concept of cultivating students’ innovative ability by combining inclusiveness with stratification, improve the cultivation mechanism of endogenous-exogenous dual empowerment, and build the cultivation mode of segmented-through mutual articulation, so that the top innovative talents can continually emerge.

Key words: students’ innovative ability, early cultivation of top innovative talents, machine learning, influencing factors