A few researchers think the estimator of randomized response may take value outside the limit [0, 1], so randomized response technique fails. In my opinion, if sample size is random and representative of the population, it will never fail.
What is your opinion?
My source of evidence is the paper: Simulated Minimum Sample Size Requirements in Various Randomized Response Models.
Please see this link:
https://www.researchgate.net/publication/260426308_Simulated_Minimum_Sample_Size_Requirements_in_Various_Randomized_Response_Models
Article Simulated Minimum Sample Size Requirements in Various Random...
Yes, I agree with you. If sample is a true representative of population then sample characteristics should be similar to that of the population. this is true under the assumption of truthful reporting of course.
This is my argument that is why we need to teach these models at high school level? If we make general public aware of these models, and once everyone will learn that they are protected by the randomization device, then they are not supposed to be liars. General public will also believe that this survey statistician is collecting information for their good future or good health, then a respondent will always be reporting truthfully. In my opinion, it is matter of developing confidence among the general public that they are protected by the randomization device, so they do not need to be liar any more.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size