数量特征随机化回答模型中的比估计和回归估计方法  

RATIO AND REGRESSION ESTIMATION PROCEDURES USED IN RANDOMIZED RESPONSE MODELS FOR QUANTITATIVE DATA

作　　者：冀鹏浩[1] 闫在在[1,2] 

机构地区：[1]内蒙古工业大学理学院,呼和浩特010051 [2]内蒙古工业大学管理学院,呼和浩特010051

出　　处：《内蒙古工业大学学报（自然科学版）》2009年第3期166-176,共11页Journal of Inner Mongolia University of Technology：Natural Science Edition

基　　金：国家自然科学基金项目(10761004);内蒙古自然科学基金项目(200607010108);内蒙古人才基金项目(2007)

摘　　要：本文研究了含辅助信息的数量特征随机化调查的理论与方法.建立和发展了数量特征随机化调查的比估计和回归估计理论与方法.对于给定的人群有限总体,构造了敏感特征指标总体均值的比估计量和回归估计量,证明了本文提出的方法优于Eichhorn BH,HayreLS(1983)和Gupta et al.(2002)提出的方法.Monte-Carlo数值模拟验证了提出方法的有效性.Investigations are conducted into the quantitative sensitive characteristic survey by use of randomized response sampling,complemented with auxiliary information.Ratio estimation and regression estimation procedures for such surveys are established and developed.Ratio and regression estimators are determined and are used to estimate the mean value of sensitive characteristic indexes of a given group of people in a confined community.It is proven that the proposed procedures have superiority over procedures suggested by Eichhorn BH,Hayre LS（1983） and Gupta et al.（2002）.The effectiveness of the proposed procedures is verified by means of Monte-Carlo numerical simulation.

关 键 词：比估计 回归估计 随机化回答 数量特征敏感变量 辅助变量 

分 类 号：O212.4[理学—概率论与数理统计]