随机化应答调查方案的参数估计问题  

Approaching parameters in the randomized response model

作　　者：章栋恩[1] 

机构地区：[1]北京工商大学数理部,北京100037

出　　处：《纯粹数学与应用数学》2000年第2期79-84,共6页Pure and Applied Mathematics

摘　　要：随机化应答调查是一种特殊的数据采集技术 ,使得调查者既得到了某个敏感问题的信息又保护了被调查者的稳私 .本文研究从两种随机化应答调查方案采得数据后的参数估计问题 .在应用Bayes估计时 ,给出了便于用Mathematica软件计算的后验均值和方差的公式 .通过模拟实验 ,比较了两个方案所得估计量的优劣 .Randomized-response technique is a very special data collecting procedure.This allows researcher to obtain sensitive information while guaranteeing privacy to respondents.In this paper,for two procedures of collecting data we research the estimators of the interesting parameters in the models.When applying Bayes inference,we propose the formulas of mean and variance of the posterior distribution which are easy to code by Mathematica.Simulating shows that the estimators in two models have different efficiency.

关 键 词：MLE估计 BAYES估计 参数估计 随机化应答调查 

分 类 号：O212.2[理学—概率论与数理统计] O211.67[理学—数学]