人口普查质量评估抽样问题探讨  

On Discussion in the Sampling of Population Census Quality Assessment

作　　者：朱喜安[1] 马樱格 Zhu Xi’an;Ma Yingge(School of Statistics and Mathematics,Zhongnan University of Economics and Law,Wuhan 430073,China)

机构地区：[1]中南财经政法大学统计与数学学院,武汉430073

出　　处：《统计与决策》2020年第20期9-14,共6页Statistics & Decision

摘　　要：中国每十年进行一次全国人口普查,人口普查质量评估主要是对人口普查的数据质量进行评估。影响数据质量的主要原因是普查中可能存在经常性的漏报、多登记和无意识的错报,我国通过构造双系统估计量进行误差估计。在以往构建的双系统估计量中,抽取样本代表性不强,没有考虑前提假定,计算估计量的方法也存在不足。文章通过改进抽样方案、调整人口普查质量评估相关工作措施和改进双系统估计量模型,提出改进办法。通过仿生数据对人口普查质量评估具体过程进行检验。China conducts a national population census every 10 years and the quality assessment of population census mainly evaluates the data quality of population census.The main reasons that affect the quality of data are the frequent omission,over-registration and unintentional misstatement in the census.In the previous two-system estimators,the sampling is not be representative enough,without taking premise assumptions into account,the method for calculating the estimator also inadequate.This paper proposed the improvement measures by improving the sampling scheme,adjusting the relevant measures of population census quality assessment and improving the dual-system estimator model,and the paper also uses bionic data to test the specific process of population census quality assessment.

关 键 词：人口普查质量评估 抽样调查方案设计 捕获-再捕获模型 双系统估计量 

分 类 号：C829.2[社会学—统计学]