摘要
在连续性抽样估计中,现有方法未考虑空间相关总体,且已有方法存在难以获得总体的辅助变量信息和未能对轮换样本估计量进行改进的不足。为解决上述问题,针对空间相关总体的连续性抽样,本文借助Kriging方法将地理坐标辅助信息引入抽样估计环节,提出基于时空二维辅助信息的回归抽样估计方法,得到相应估计量并推导估计量方差,同时对估计量的无偏性进行证明。该方法考虑了空间相关总体,提供了一种在连续性抽样估计中利用地理坐标辅助信息的思路,规避了对总体辅助信息的获取。并且同时适用于重复样本调查和轮换样本调查,在轮换样本调查中能同时对轮换样本估计量和非轮换样本估计量进行改进。随后通过数值模拟,验证了该方法同时适用于均匀分布、随机分布和聚集分布三种常见空间分布类型,且随着总体空间相关性、样本量和样本轮换比率的提高,本文提出的估计量相较于Horvitz-Thompson估计量和传统组合估计量的估计优势不断增强,显示了该估计量的优良性能。最后通过实证分析,验证了该方法在实际应用中的可行性,为后续在抽样估计环节考虑空间相关总体和有效利用地理坐标辅助信息提供了理论基础和研究参考。
In successive sampling estimation,the existing methods do not consider the spatial correlation population,and have the defects of difficult to obtain the auxiliary variables'information of the popula-tion and failing to optimize the rotating portion estimator.To solve these problems,for the successive sampling in spatial correlation population,this article take advantage of the Kriging method to introduce the auxiliary information of geographic coordinates into the sampling estimation process,proposes the regression sampling estimation method based on two-dimensional spatiotemporal auxiliary information,obtains the corresponding estimator and estimator variance,and also proves the unbiasedness of the estimator.This method takes the spatial correlation population into consideration and provides an idea of using the auxiliary information of geographic coordinates in successive sampling estimation,which avoids the acquisition of the population auxiliary information.Besides,it's suitable for both repeated survey and rotating panel survey,and can also optimize the rotating and non-rotating portion estimators simultaneously in rotating panel survey.Then through numerical simulation,it is verified that the method is suitable for three common types of spatial distribution:uniform distribution,random distribution and clustered distribution.What's more,with the increase of population spatial correlation,sample size and rotating rate,the estimation advantage of the estimator proposed in this paper has continued to increase compared with the Horvitz-Thompson estimator and traditional combination estimator,which showing the excellent performance of the estimator.Finally,through empirical analysis,the feasibility of the method in practical application is tested,which providing theoretical basis and research references for the follow-up research on considering the spatial correlation population and introducing the geographic coordinate auxiliary information into the sampling estimation process.
作者
杨颖琪
陈光慧
YANG Ying-qi;CHEN Guang-hui(School of Economics,Jinan University,Guangzhou 510632,China)
出处
《数理统计与管理》
北大核心
2025年第4期589-606,共18页
Journal of Applied Statistics and Management
基金
国家社会科学基金一般项目(18BTJ005)
全国统计科学研究重大项目(2024LD001)。
关键词
空间相关总体
地理坐标辅助信息
时空二维辅助信息
回归抽样估计
KRIGING方法
spatial correlation population
the auxiliary information of geographic coordinates
two-dimensional spatiotemporal auxiliary information
regression sampling estimation
Kriging method
