统计研究 ›› 2007, Vol. 24 ›› Issue (8): 68-71.

• 论文 • 上一篇    下一篇

n>2时πps抽样下方差估计量的算法实现

陈光慧 刘建平

  

  1. 暨南大学经济学院统计学系
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2007-08-15 发布日期:2007-08-15

A Realization of Computing Method of Variance Estimator under Sampling when

Chen Guanghui ;Liu Jianping   

  • Received:1900-01-01 Revised:1900-01-01 Online:2007-08-15 Published:2007-08-15

摘要: 为了解决 抽样下方差估计量计算困难的难题,本文引入了Hájek近似方法,给出了二阶包含概率的一个近似公式。在此基础上,推导出Hájek近似方差,并根据Hájek近似方差公式提出了一个Hájek近似方差的估计量。通过实际数据的分析进一步表明,经过Hájek近似的方差估计量在样本量大于2时,不仅计算简便,而且估计准确,与真实的方差估计量相差甚小。

关键词: 抽样, Horvitz-Thompson估计量, Hájek近似

Abstract: In order to solve the problem of computing complex variance estimator for sampling, this paper uses Hájek approximation approach to obtain the approximation of joint inclusion probability, Hájek variance approximation and the estimator of Hájek variance.Using this variance estimator to do the empirical research, we draw the conclusion that sampling is easily implemented when the size of sample is larger than two because of the simplicity of computing variance.


 

Key words: sampling, Horvitz-Thompson estimator, Hájek approximation