摘要
源于教育的健康效应研究中存在相互冲突的研究结论的现象,本文对中英文期刊发表的中国教育的健康效应的实证研究文献进行了Meta分析。FAT-PET检验和PEESE检验表明,已发表的研究文献中存在着显著的发表偏误,在剔除发表偏误后,中国教育对健康的真实影响效应接近于零;多元Meta回归模型的Post Lasso估计量和Double Post Lasso估计量表明,健康指标选取、教育指标选取以及是否解决内生性是导致研究结论差异的三个重要来源。
Originating from the phenomenon of conflicting research conclusions in the research on the health effects of education,this paper conducts a meta-analysis of empirical studies on the health effects of education in China.The FAT-PET test and the PEESE test show that there are significant publication errors in the published research literature.After excluding publication errors,the true effect of education on health in China is close to zero.The Post Lasso estimators and Double Post Lasso estimators of multiple meta-regression models show that the selection of health indicators,the selection of education indicators and whether to solve endogeneity are three important sources leading to the differences in research conclusions.
作者
彭树宏
PENG Shu-hong(School of Public Finance and Public Administration,Jiangxi University of Finance and Economics)
出处
《教育经济评论》
CSSCI
2022年第1期99-120,共22页
China Economics of Education Review
基金
江西省教育科学规划课题“大学教育对健康的影响效应及其机制研究”(19YB051)。
作者简介
彭树宏,江西财经大学财税与公共管理学院,电子邮箱地址:shuhong.peng@gmail.com。
