分层整群抽样的Warner模型RRT技术及其对大学生婚前性行为调查中的应用  

Stratified Cluster Sampling and Its Application on Warner Model for Sensitive Question Survey

作　　者：高歌[1] 范玉波[1] 

机构地区：[1]苏州大学公共卫生学院,江苏苏州215123

出　　处：《数理统计与管理》2010年第1期185-190,共6页Journal of Applied Statistics and Management

基　　金：国家自然科学基金资助项目(编号30972548)

摘　　要：本文的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。使用Cochran W.G.的抽样理论、随机应答技术的Warner模型、全概率公式、方差的基本性质等理论与方法,推导出二分类敏感问题随机应答技术Warner模型在整群抽样、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信度较高的成功应用效果。To explore scientific sampling methods and corresponding formulas for sensitive question survey on stratified cluster sampling . Cochran W.G.＇s classic sampling theories, Warner model, total probability formulas and properties of variance were used in this paper. Formulas for the estimate of the population proportion and its variance on Warner model in cluster sampling and stratified cluster sampling were deduced. With high reliability, our survey methods on Warner model may have an extensive application in sensitive question survey.

关 键 词：敏感问题 WARNER模型 分层整群抽样 总体比例的估计 估计方差 

分 类 号：O212[理学—概率论与数理统计]