期刊文献+

学生特征、学校特征与学生学业表现——基于机器学习方法的实证研究 被引量:1

Student Characteristics,School Characteristics and Student Academic Performance:An Empirical Study Based on Machine Learning Approach
下载PDF
导出
摘要 基于PISA2018我国四省市学生和学校的调查数据,在教育生产函数的框架下,通过机器学习的研究方法,探讨不同学生和学校特征在学业表现中的重要性。研究发现,在学生特征方面:个体社会经济地位和课上学习时间在数学、阅读、科学表现中的重要性最突出,且学习时长对学业成绩的影响存在学科差异;学生职业期望在数学和科学成绩上的重要性较为突出;学生的社会经济地位及课上学习时长在高、低学业表现学生群体中的重要性均相对突出。在学校特征方面,弱势学生比例及教师数量在三种学业表现中都相对重要,且其在高、低学业表现学校群体中的重要性相对突出。此外,比较发现,在初中阶段,学校特征类因素对学生学业表现的影响更大。因此,建议提高学生课堂学习的效率,增强青少年对基础学科的认同感;推进公共教育服务均等化,加大对教育资源优越地区相对弱势学生群体的关注力度;调整初中阶段教育资源配置结构,优先保障对学校的投入。 The current study used PISA 2018’s data on Beijing,Shanghai,Zhejiang,and Jiangsu,based on the theoretical framework of educational production function,and used the machine learning analysis approaches to explore the importance of various student and school characteristics in students’academic performance.The results showed that,regarding student characteristics,students’SES and in-class study time were the most important characteristics associated with students’math,reading,and science performance,and in-class study time showed differentiate effect across subjects;student’s expected occupational status was important for math and science performance but not for reading;the common group characteristics that shared by the highest and lowest academic performance group were students’SES and in-class study time.Regarding school characteristics,disadvantage student ratio and the total number of all teachers at school were the relatively important school characteristics for students’academic performance,and they are the common group characteristics that shared by the highest and lowest academic performance group.In addition,the current study revealed that school characteristics were more important than student characteristics for students’academic performance in secondary education.Therefore,we recommended that improving students’in-class study efficiency,and enhancing their identity towards basic disciplines;accelerating the equalization of public educational services,and raising greater attention to relative disadvantage students;adjusting educational resource allocation in secondary education,and prioritizing the support for school improvement.
作者 赵宇阳 陈越洋 桑标 Yuyang Zhao;Yueyang Chen;Biao Sang(School of Sociology,Shanghai University,Shanghai,200444;Lab for Educational Big Data and Policymaking,Shanghai Academy of Educational Sciences,Shanghai,200032)
出处 《教育与经济》 CSSCI 北大核心 2023年第1期47-58,共12页 Education & Economy
基金 教育部“教育大数据与教育决策实验室”2022年度重点资助项目 中国社会科学院-上海市人民政府上海研究院资助。
关键词 PISA2018 教育生产函数 机器学习 决策树回归 随机森林回归 PISA 2018 education production function machine learning decision tree random forest
  • 相关文献

参考文献16

二级参考文献198

共引文献353

同被引文献19

引证文献1

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部