三维平衡轮换的多水平抽样调查设计研究综述及应用  

The Design Study Review and Application of Multi-Level Sampling Survey Based on Three-Way Balanced Rotation Scheme

作　　者：陈光慧[1] 

机构地区：[1]暨南大学经济学院,广东广州510632

出　　处：《统计与信息论坛》2013年第11期9-15,共7页Journal of Statistics and Information

基　　金：教育部人文社会科学研究青年基金项目<模型辅助抽样估计方法及其在我国的应用研究>(13YJC910002);中央高校基本科研业务费专项资金(暨南启明星计划)项目<我国政府统计调查与数据调整分析研究>(12JNQM006)

摘　　要：现行的多水平抽样调查使用的各种形式的轮换模式,在西方各国均得到了广泛应用,但也存在着一系列问题。鉴此,通过对各种形式轮换模式的归纳统一和理论化综述研究,最终归纳出三维平衡多水平轮换模式设计方法,即将多水平轮换模式设计与后续的抽样估计方法研究统一起来,不但能够削减各类轮换偏差的负面影响,还能准确度量轮换样本之间的相关关系,并在多水平调查下得出更加准确的连续性抽样数据。此套设计方法具有极大的推广价值。The current multi-level sample surveys use various types of multi-level rotation schemes which many of western countries have been widely used. However, there are also a number of issues. Therefore, various types of rotation schemes are unified and the systematic, theory-based research is carried out. Then the scheme design of three-way balanced multi-level rotation is concluded, and the advantages of its application are summarized. This design approach not only unifies multi-level rotation scheme design and the following study of the sampling estimation method, but also reduces the negative impact of various types of rotation biases and measures the correlation between the rotation samples accurately. Ultimately, more accurate successive sampling data from multi-level survey is obtained.

关 键 词：三维平衡轮换 多水平轮换 模式设计 轮换偏差 

分 类 号：C811[社会学—统计学]