统计研究

• 论文 • 上一篇    下一篇

PPS抽样下匹配系数的构建及应用

艾小青 胡丹丹   

  • 出版日期:2014-12-15 发布日期:2014-12-15

Design and Apply Matching Coefficient in PPS Sampling

Ai Xiaoqing Hu Dandan   

  • Online:2014-12-15 Published:2014-12-15

摘要: PPS抽样时,一般认为当规模变量与目标变量的相关性越强,抽样的精度越高,但少有文献研究相关性的具体内涵以及与抽样精度的内在关联。本文提出构建匹配系数反映规模变量与目标变量的比率相关关系,并对PPS抽样下的样本匹配系数进行了构建和分析,还通过Monte Carlo模拟揭示了该统计量的良好性质。本文最后揭示了样本匹配系数在抽样精度评价时有着良好的应用效果。

关键词: PPS抽样, 规模变量, 匹配系数, 方差估计

Abstract: It is generally believed that in PPS sampling, when the auxiliary variable has more correlation with the target variable, the sampling precision will be higher, but there is little research on specific connotation of correlation coefficient and its intrinsic connection with the sampling precision. In this paper, matching coefficient is proposed to reflect the ratio correlation between the auxiliary variable and target variable, then sample matching coefficient based on PPS sampling is built and analyzed, and it shows the good properties of the statistics by Monte Carlo simulation. Finally, this paper reveals that the sample matching coefficient has good application effect in sampling precision evaluation.

Key words: PPS sampling, auxiliary variable, matching coefficient, variance estimation