我国各地区教育发展水平的无交叉分位回归模型

Noncrossing Quantile Regression Modelling for Regional Education Development Data in China

摘要: 本文基于无交叉分位回归方法对我国各地区教育水平发展现状进行分析，研究了不同地区人口的平均受教育年限与人均GDP、教育经费和教师资源之间的关系。由于我国各地区教育水平发展不均衡，分位回归方法能够反映出数据的全貌，而无交叉分位回归能够使参数估计更加合理。研究结果表明，在教育水平较低的地区，教育发展受到经济条件的制约，在教育水平居中的地区，教育发展更多地受到教师资源的制约，而教育经费对不同地区的影响较为复杂。

Abstract: In this paper, we study regional education development in China based on noncrossing quantile regression and focus on the relationship between average years of education and average GDP, education budget and teacher resources. On account of the imbalance in regional education development, quantile regression offers a complete picture of data; on the other hand, noncrossing quantile regression makes the estimators more reasonable. The results prove that economic condition hinders the development of education in low level region, while in middle level region teacher resources play a more important role, and the effects of education fund on different regions are more complex.