Estimation of parameters in stratified cluster sampling on randomized response technique for sensitive question survey
10.3724/SP.J.1008.2009.00170
- Author:
Ge GAO
1
Author Information
1. Department of Health Statistics
- Publication Type:Journal Article
- Keywords:
Estimated variance;
Estimation of parameters;
RRT model;
Sensitive question;
Stratified cluster sampling
- From:
Academic Journal of Second Military Medical University
2010;30(2):170-177
- CountryChina
- Language:Chinese
-
Abstract:
Objective: To explore a scientific, complicated sampling method and corresponding formulas for sensitive question survey. Methods: Cochran W. G.'s classic sampling theories, two randomized response technique(RRT) models, total probability formulas, and properties of mean and variance were used in this study. Results: Formulas for the estimation of the population proportion and its variance for dichotomous sensitive questions on Greenberg model in cluster sampling and stratified cluster sampling were deduced. Formulas for the estimation of the population mean and its variance for quantitative sensitive question on the additive constant model in cluster sampling and stratified cluster sampling were deduced. Our survey methods and formulas were successfully applied in surveying the pre-marriage sex and cheating in examination in Soochow University. Conclusion: Our survey methods and formulas on two RRT models for sensitive question survey in cluster sampling and stratified cluster sampling have high reliablility.
