中文摘要：

为了评估2002年人口与住房普查质量,乌干达首次进行事后计数调查（Post Enumeration Survey,简称PES）。为了避免重叠,PES在普查结束后进行。本质上来看,PES属于抽样调查,以计数区为抽样单位,使用分层、单阶段、与规模成比例的抽样方法抽取样本。使用问卷采集计数区家庭、个人信息,并进行比较。对未匹配家庭或个人现场调整。利用抽样权数和比较数据计算总体指标,在此基础上,估计普查涵盖误差和内容误差。本文给出了总体总数估计量及其抽样方差估计量,以及内容误差的区间估计量。最后总结了乌干达这次PES教训,并对其未来PES改进提出了具体的建议。

英文摘要：

Uganda carried its first post enumeration survey（short for PES） to evaluate its population and housing census of the year 2002.PES was carried after the census to avoid being overlapped.Basically,PES was a sample survey,The sampling unit was enumeration area,and the samples were systematically drawn through a one-stage stratified cluster with probability proportional to measures of size.Questionnaires were used to collect households or persons from enumeration areas and then compared.Field reconciliation was carried for non-match households or persons.Population counts was estimated according to sampling weights and sample data,based on which,coverage and content errors were estimated.This article gave the estimators and their sample variance estimator of population counts and interval estimator of content errors.Finally,the PES lessons was outlined and suggestions were put forward for the future PES.