双系统估计模型以往在美国人口普查涵盖测量调查中主要用来测量净误差,美国2010年人口普查涵盖测量中考虑依据Mulry和Kostanich提出的误差框架测量普查涵盖误差的构成。本文分析了双系统估计模型理论与其在人口普查实践应用中的不足,通过增加两项模型基本假设,结合前人研究,提出构建适用于一般普查类型的扩展双系统估计模型,通过证明其两条重要的匹配性质,验证了有关学者的研究假设;以此为基础,结合现有数据条件,讨论了基于双系统估计模型原理的普查涵盖误差测量应用。
U.S. Census Bureau's previous coverage measurement surveys were designed primarily to estimate net Census Coverage Errors (CCE) using Dual System Estimation Model (DSEM). A recom- mendation on developing plans for measuring the components of CCE was given in U.S. 2010 census coverage measurement using a framework provided by Mulry and Kostanich (2006). Firstly, this paper analysed the lacks between DSEM in theory and application. Two assumptions were added to the model. Secondly, expanded DSEM, to measure a general census coverage error in China, was constructed based on ideas of the framework and the added assumptions. At last, matching properties were proved and an application research under the conditions of the existing available data was also discussed in this process.