基于双重抽样框的估计方法评述与启示

The review and implication about estimation methods based on dual sampling frame

在经济和信息高速发展的当今社会,抽样调查中被调查对象的流动日益频繁,传统的单一抽样框已很难完整覆盖流动性较强的目标总体。抽样实践中,通常采用两个不完整抽样框的组合可以实现对目标总体的完整覆盖。基于双重抽样框进行抽样调查时,由于样本在两个抽样框中存在交叉,致使抽样估计甚是困难。目前国外已有的各种双重抽样框估计方法,可分为分离抽样框估计和组合抽样框估计两类,前者的不足是要求所有样本单位都能被正确分组,后者的弱点是需事先知道重叠部分在两个抽样框中的包含概率,这些通常不易实现。针对这些不足进一步改进的思路是:保持估计量的稳定性,允许错误归类的情况发生,尽可能使计算容易、操作方便。

With the rapid development of economic and information,the objects in the sampling survey flowing increasingly frequent.Traditional single sampling frame is difficult to cover the overall goal its mobility is relatively strong.In the sampling practice,usually using a combination of two incomplete sampling frames is completely able to cover the population.If we carry on the sample investigation based on the dual sampling frames,its sampling estimation is very difficulty because of the sample exists in two sampling frames overlapping.Systematically reviews present overseas existing dual sampling frame estimation methods,and divides these methods into the separation sampling frame estimation and the combination sampling frame estimation,and commentaries the various of methods' disadvantage to give further improve.To provide the instruction about adoption the dual sampling frame investigations for our country.