条件配置抽样的抽样设计

Sampling Design of Conditional Collocated Sampling

文章提出一种新的称之为条件配置抽样的抽样设计:通过不断产生配置样本,直到其样本量实现了预定容量n时样本才被接受。一方面围绕建立该设计基本理论展开探讨研究,给出了条件配置抽样一阶、二阶包含概率的形式表达式;另一方面在经典的线性超总体模型下,从数值上说明与条件Possion抽样比较,条件配置抽样在精度和实施上的优良性。

This paper presents a new sampling design called conditional collocated sampling: By continuously generating collocated sampling, the samples are not accepted until their sample sizes achieve a predetermined capacity n. On the one hand,based on the establishment of the basic theory of the design, this paper makes a discussion and study to offer the formal expressions of the first-order and second-order inclusion probability of conditional configuration sampling. On the other, in the classical linear super-population model, it is numerically illustrated that conditional allocation sampling is superior in accuracy and implementation compared with conditional Possion sampling.