中国人口受教育情况统计的多目标抽样设计  

Multi-Objective Sampling Design of Chinese Population Education Statistics

作　　者：马金萍[1] 任慧[1,2] 

机构地区：[1]西安财经学院统计学院,西安710100 [2]郑州财经学院信息工程学院,郑州450049

出　　处：《统计与决策》2017年第13期5-8,共4页Statistics & Decision

基　　金：国家哲学社会科学基金资助项目(13BJ006);国家统计科研重点项目(2012LZ020)

摘　　要：文章在考虑中国每十万人口中目前在校生情况和抽样调查可行性基础上,提出了在校生受教育人口分层三阶段抽样。第一阶段首先根据在校生总人口进行聚类分析,而后进行分层,第二、三阶段采取随机起点系统抽样或者PPS抽样。在全国近几年样本中的样本量已调查的条件下,对抽取的所有样本进行事后分层和适当追加样本量,对各省之间人口受教育程度人数所占人口比例进行估计,同时可满足全国、部分高校、部分城市等多目标要求的人口受教育情况的目标量估计,进而可以了解各省大多数人受教育层次。In consideration of the current number of the students at school per one hundred thousand Chinese people and the feasibility of the sampling survey, this paper puts forward a hierarchical three-stage sampling method for the students at school. In the first stage, the paper makes a cluster analysis for the total number of the internal students before doing stratification. In the sec- ond and third stage, the random starting-point system sampling or PPS sampling is used. Following the investigation of the sam- ples for recent years in China, the paper makes a post-stratification for all the samples and properly enlarges the sample size to es- timate the proportion of the population＇ s educational level among the provinces. At the same time, the proposed method is able to meet the muhi-objective requirements for estimation of the population＇ s educational level of some colleges, some cities or even the whole country. Moreover it helps to learn the educational level of the most of the people in every province in China.

关 键 词：人口受教育情况 多目标抽样设计 聚类分析 PPS抽样 

分 类 号：O212[理学—概率论与数理统计]