抽样调查中的校准估计法  

The Calibration Estimation Approach In Sampling Survey

作　　者：刘建平[1] 常启辉 

机构地区：[1]暨南大学经济学院统计学系 [2]暨南大学经济学院

出　　处：《统计研究》2014年第12期92-100,共9页Statistical Research

基　　金：国家社科基金项目"我国政府统计调查体系研究"(12ATJ002);教育部人文社科基金项目"改革和完善我国政府统计调查体系研究"(11YJA910006)的阶段性成果

摘　　要：本文梳理总结了校准估计法自首次提出以来的研究成果。理论方法的发展集中在最短距离法、工具向量法和模型校准法三种校准方法的研究上;方法应用的发展体现在对简单参数和复杂参数的校准估计上,特别对小域估计、无回答、二重抽样等特定抽样问题和总体分位数、总体方差估计中校准估计法的具体应用做了重点梳理介绍。最后,本文对校准估计法的理论和应用研究前景做了展望。This paper summarizes the achievements of calibration estimation approach since it was first proposed. The development of theories and methods focuses on the study of three calibration approaches ,including the minimum distance method ,the instrument vector method and the model-calibration method. The development of its application is embodied in the calibration estimation of simple parameters and complex parameters. This paper puts great emphasis on the introduction of specific survey problems, such as small domain estimation , nonresponse and two-phase sampling , as well as the application of calibration estimation for population quantiles and variance , and gives an outlook of the theoretical and applied research of calibration estimation approach.

关 键 词：校准估计量 辅助信息 权数 

分 类 号：C811[社会学—统计学]