Enhancing a Randomized Response Model to Estimate Population Means to Sensitive Questions

被引:4
|
作者
Dihidar, Kajal [1 ]
Chowdhury, Joydeep [2 ]
机构
[1] Indian Stat Inst, Sampling & Official Stat Unit, Kolkata 700108, W Bengal, India
[2] Indian Stat Inst, Bayesian & Interdisciplinary Res Unit, Kolkata 700108, W Bengal, India
关键词
efficiency in estimation; scrambling variable; sensitive attribute; unequal probability sampling; COMPLEX SURVEY; EFFICIENCY; DESIGNS;
D O I
10.1080/08898480.2013.816216
中图分类号
C921 [人口统计学];
学科分类号
摘要
The population means of variables such as expenditure on alcohol, abortion, amount of dowry are estimated on modifying Gjestvang and Singh (2009) randomized response model. Instead of simple random sampling with replacement scheme, the respondents are allowed to be chosen by any varying probability sampling scheme, which in particular, also includes simple random sampling with replacement scheme. Whatever the sampling design, the modified estimators perform better than the usual estimator.
引用
下载
收藏
页码:123 / 136
页数:14
相关论文
共 50 条