不等概率抽样估计的原理与应用

The Theory of Inequiprobability Sample Estimation and Its Applications

不等概率抽样估计是一种十分有效的抽样推断方法.它在以样本估计总体时是应用一个与所要调查的目标变量相关的辅助变量作为样本单位抽选的概率,能使得对总体指标数值影响大的单位赋予大的抽中概率,同时在构造样本指标时,可利用辅助变量值对目标变量值进行调整,增大了样本对总体的代表性,由此抽取的样本和依据样本计算的样本指标就更接近于总体指标.

Inequiprobability sample estimation is a very effective method for sample inference. When estimation population by sample, the method uses the probability of an auxiliary variable related to the target variable as the probability of selecting the sample in order to give greater weight to the sample points that effect the population greater, meanwhile, when constructing indexes, the method magnifies the representativeness of sample by adjusting the value of target variable according to the value of auxiliary variable, so the sample selected in this way and the value of target variable got from these sample points will be closer to the population indexes.