论文部分内容阅读
本次研究旨在通过分层整群随机抽样抽取一个经济而有效的国家卫生统计和专题调查的地区样本。采用人口普查分地区社会经济、文化教育和健康等多个指标的资料,用主成分因子分析法确定抽样分层的标识,在此基础上用K-MEANS聚类分析法对总体聚类分层;采用分层整群随机抽样技术抽取多组不同样本容量的样本,分别计算出每个样本各变量的统计量;通过各样本变量的统计量对代表总体参数精确度的比较和各样本变量的统计量与总体参数分布拟合度检验确定最佳样本容量和样本地区。结果表明,抽取60个县(市)可以代表全国,至少90个县(市)可以代表全国不同类型地区。
This study aims to extract an economical and effective regional sample of national health statistics and special surveys through stratified cluster random sampling. Using the data of multiple indicators such as socio-economic, cultural education, and health in the census, we used the principal component analysis to determine the stratified label, and based on this, we used K-MEANS clustering to stratify the population. The use of stratified cluster random sampling techniques to extract multiple samples of different sample sizes, calculate the statistics of each sample variable; through the statistics of each sample variable on behalf of the overall parameters of the accuracy of the comparison and the various sample variables The test of fitting the statistics with the overall parameter distribution determines the best sample size and sample area. The results show that 60 counties (cities) can be selected to represent the whole country, and at least 90 counties (cities) can represent different types of regions across the country.