Sample Surveys Of Sensitive Issues Basic And Theoretical Research

The randomized response technique (RRT) is originally developed for qualitative purpose of sensitive sampling survey. Since Warner and Simmons developed Warner model and Simmons model applying randomized response technique, many statistic worker have developed many new modified model according to the two models. In this paper, the author introduced the two important models and some modified models. Since the idea of one of the modified model named two sensitive questions Warner- Simmons union model can not dissolve multi-choice sensitive question sampling survey, we develop a new idea. Through the new idea, we can dissolve this question easily. Another, although the protection of the broadly-defined Simmons model is very good, and the model is more efficient than Greenberg model and Moor model, we discover that some little modify can make the model more efficient.The research of sensitive questions bases on real responses of the surveyed people .But in fact not real responses is ordinary in sensitive questions sampling survey. This not real response will make our estimator inefficient. So the research of sensitive questions sampling survey with not real responses is essential. Since the research of these questions is more complex than real responses, the paper about this research is very few. Through the research of these questions we develop a new method on this situation. When surveyed people response sensitive question randomly, this model is efficient. But in another situation, the new model is not efficient. In this situation we need deeply research, at the same time the research is important and essential.