摘要
地理信息系统的广泛使用为抽样调查中引入了总体单元的空间信息,随之产生的空间相关性破坏了总体单元之间的独立性假设,传统抽样方法在空间总体中的应用也面临着样本代表性下降的困扰。针对这一问题,提出了利用地理坐标信息选取空间平衡样本以提高样本代表性,并利用样本中包含的地理坐标信息改进方差估计量以提高估计效率。模拟研究和实证分析结果表明,基于空间平衡样本的统计推断在空间相关性较强时能够显著提高估计效率。
Spatial dependency in surveys breaks the assumption of independence in traditional sampling methods.Spatial balanced sample is used to improve the representative of sample and spatial auxiliary information is used to improve the variance estimator in order to gain more efficiency.It is shown by simulation and case study that statistical inference based on spatial balanced sample gains more efficiency of variance estimator under a strong spatial dependency in the population.
作者
郝一炜
金勇进
HAO Yi-wei;JIN Yong-jin(Beijing Ditan Hospital Capital Medical University,Beijing 100015,China;Applied Statistical Science Research Centre in Renmin University of China,Beijing 100872,China;School of Statistics in Renmin University of China,Beijing 100872.China;Beijing Municipal Health Commission Medical Management Data Quality Control and Iniprovement Center,Beijing 100035,China)
出处
《数理统计与管理》
CSSCI
北大核心
2020年第6期978-989,共12页
Journal of Applied Statistics and Management
基金
国家社科基金项目(15BTJ014)
中国人民大学双一流建设项目。
关键词
地理坐标
空间信息
空间平衡样本
空间相关性
geographic coordinate information
spatial information
spatial balanced sample
spatial dependency
