人口普查涵盖误差的估计  

Estimation of Population Census Coverage Error

作　　者：胡桂华[1] 

机构地区：[1]广西财经学院数学与统计系,广西南宁530003

出　　处：《井冈山大学学报（社会科学版）》2011年第4期69-75,共7页Journal of Jinggangshan University(Social Sciences)

基　　金：国家社会科学基金项目"我国人口普查质量评估方法研究"(项目编号:L10XTJ003);全国统计科学重大项目"行政记录在人口与就业统计中的应用"(项目编号:2009LD003);广西自然科学基金资助项目"人口普查涵盖误差的估计"2010GXNSFA013107;广西人文社会科学发展研究中心科学研究工程2010年度开放基金资助项目"广西人口数目估计方法研究"(项目编号:KF2010021)~~

摘　　要：人口普查不可避免产生误差,包括涵盖误差和内容误差。涵盖误差可进一步分为净涵盖误差及遗漏和错误计数。一般是先估计净涵盖误差和错误计数,然后估计遗漏。净涵盖误差为真实人口数与普查人口数之差。目前世界上大多数国家是使用人口普查后的事后计数调查来估计真实人口数。事后计数调包括抽样设计、调查样本、比较样本、事后分层和基于捕获-再捕获模型的双系统估计量的构造。我国每次人口普查后都进行了事后计数调查和估计了净涵盖误差,但我国的做法与其他国家相比存在某些缺陷,需要改进与完善。本文对我国事后计数调查提出若干对策建议。Population census contains inevitable errors, including coverage error and content error.Coverage error is further classified into net coverage error, erroneous enumerations and omissions. Generally, net coverage error and erroneous enumerations are first estimated and then omissions are estimated. Net coverage error is difference between true and census population count. At present most countries use the post- enumeration survey after census to estimate true population count. The post-enumeration survey includes sample design, sample survey and comparison, post-stratification, and dual system estimator construction based on capture-recapture model. China carried out post- enumeration survey after each census and estimate net coverage error. However Compared with methods used by other countries the Chinese one is beset with some drawbacks and need improvement. This paper puts forwards some advices on post-enumeration survey for China.

关 键 词：人口普查 人口普查的事后计数调查 捕获-再捕获模型 涵盖误差 

分 类 号：C921.2[社会学—人口学]