本文提出一种同时估计个体报告真实回答的概率T和敏感指标均值μx的无关问题模型,文中讨论了T和μx估计量的性质以及样本容量的匹配。本文计算了所提出方法与直接调查方法的相对效率,结论表明,存在不真实回答情况下,本文提出方法可以比直接调查更有效。
In this paper, we estimate simultaneously the mean μx and the probability of truthful answer T using a two stage unrelated question model. We discuss the advantages of using the proposed method under considering incompletely truthful answers of respondents.