摘要
文章以SAR模型、SEM模型和SDM模型为基础,讨论分析了未完全考虑空间因素、通漏重要解释变量、选择了错误的模型结构等常见的空间回归模型设定偏误的后果,并针对如何在实际应用中设定出正确完善的空间回归模型,提出了一些基于建模实践的设定空间回归模型的思路与方法,这些思路与方法通过实际经济问题和数理推演得以验证。
Based on the SAR model, the SEM model and the SDM model, this paper makes a discussion and analysis on such frequent consequences from specification errors of spatial regression models as no full consideration of spatial factors, missing im- portant explained variables, selection of false models and son on. The paper also aims at how to specify a right and perfect spatial regression model to propose some specification ideas of spatial regression models based on the modeling practice. These ideas and approaches are verified through the actual economic problems and mathematical derivation.
出处
《统计与决策》
CSSCI
北大核心
2018年第1期5-9,共5页
Statistics & Decision
基金
国家社会科学基金资助项目(16CTJ009)
关键词
空间回归模型
模型设定
模型结构
spatial regression model
model specification
model structure