沃纳模型的研究及优化

在敏感性问题的随机化调查中,由于涉及到被调查者的隐私问题,很难得到被调查者的配合,因此造成了调查的困难、调查结果的不准确.解决该类问题的一般方法是敏感性问题的随机化回答调查法,即沃纳模型.沃纳模型是以被调查者的回答都是真实的为前提条件的,然而事实并非如此.本文在原模型中引入了先验概率来优化沃纳模型,进而又给出了对先验概率进行估计的可行方法.分析表明,优化后的模型与现实情况较为吻合,调查结果更为准确.

Research and optimization of Warner Model

When randomized survey is made of sensitive questions,informant's privacy is dealt with,so it's difficult for informant to cooperate with investigation.It results in the difficulty of investigate and wrong result.Solution to this problem is randomized response investigate,that is,Warner Model.The correct answer of informant is a condition of Warner Model,but the condition doesn't accord with fact.In the paper transcendental probability is brought in to optimize the Warner Model,then a feasible method with w...