2019/12/12 by Zakry Zapata, Stephen A. Sedory, Sarjinder Singh
Mathematics · #Survey Sampling and Estimation Techniques #Statistical Methods and Bayesian Inference #Census and Population Estimation
paper · doi:10.1177/0049124119882469
In this article, we consider the use of the zero-truncated binomial distribution as a randomization device while estimating the population proportion of a sensitive characteristic. The resultant new estimator based on the zero-truncated binomial distribution is then compared to its competitors from both the efficiency and the protection point of views. The situations where the proposed new estimator performs better are reported. R codes are also included.