中文摘要：

双系统估计量是目前估计目标总体真实人口数及人口普查净误差的主流方法，构造双系统估计量时，要求总体中的人口具有相同的登记概率。为达此目的，目前各国通行的做法是对人口总体进行抽样后分层。用Logistic回归模型取代抽样后分层是人口普查质量评估领域的前沿问题。本文系统地解读了基于Logistic回归模型的双系统估计量及其方差估计量，认为Logistic回归模型能够纳入更多的分层变量，具有很好的应用前景。

英文摘要：

Dual system estimator is now a main method that estimates true number of persons of the target population and census net error. Constructing dual system estimator de- mands persons of population to own the same registration probability. The current practice of countries is stratifying population after sampling In order to realize this aim. Logistic regres sion model can be used instead of stratifying after sampling, which is an international fron tier issue in the field of the census quality assessment. Dual system estimator and its variance estimator is roundly and systematically interpretated by this thesis. The results of the study show that logistic regression model can include good application prospect. more post stratification variables, and has good application prospect.