敏感问题改进的随机应答技术模型分层整群抽样研究及应用  

Stratified Cluster Sampling and Its Application on the Improved RRT Model for Sensitive Question Survey

作　　者：高歌[1] 范玉波[1] 

机构地区：[1]苏州大学放射医学与公共卫生学院,江苏苏州215123

出　　处：《苏州大学学报（医学版）》2008年第5期750-754,共5页Suzhou University Journal of Medical Science

基　　金：国家自然科学基金资助项目(30571620)

摘　　要：目的为敏感性问题提供科学的较复杂抽样调查方法及其统计量的计算公式。方法使用Cochran的经典抽样理论、改进的随机应答技术、全概率公式、方差的基本性质等理论与方法进行公式推导。结果推导出二分类敏感问题改进的随机应答技术在整群抽样、分层整群抽样下总体比例的估计量及其估计方差的计算公式,并在苏州大学学生婚前性行为的调查中取得了成功的应用效果。结论该文提供的敏感性问题改进随机应答技术的整群抽样、分层整群抽样方法信度较高。Objective To explore scientific sampling methods and corresponding formulas for sensitive question survey on the complicated sample methods. Method Cochran＇s classic sampling theories, the improved randomized response technique（RRT） model, total probability formulas and properties of variance were used in this paper. Results Formulas for the estimation of the population proportion and its variance on the improved RRT model in cluster sampling and stratified cluster sampling were deduced. Our survey methods and formulas on the improved RRT model may have an extensive application for sensitive question survey at Suzhou University. Conclusion Our survey methods and formulas on the improved RRT model in cluster sampling and stratified cluster sampling are reliable.

关 键 词：敏感问题 随机应答技术 分层整群抽样 总体比例 估计方差 

分 类 号：R195.1[医药卫生—卫生统计学]