Abstract
In this paper, we propose an empirical log-likelihood estimator for estimating the population mean of a sensitive variable in the presence of an auxiliary variable. A new concept of conditional mean squared error of the empirical likelihood estimator is introduced. The proposed method is valid for simple random and without replacement sampling (SRSWOR) and could easily be extended for complex survey designs. The relative efficiency of the proposed pseudo-empirical log-likelihood estimator with respect to the usual, and to a recent estimator due to Diana and Perri (2009b), has been investigated through a simulation study.
Original language | English (US) |
---|---|
Pages (from-to) | 345-351 |
Number of pages | 7 |
Journal | Statistics and Probability Letters |
Volume | 81 |
Issue number | 3 |
DOIs | |
State | Published - Mar 2011 |
Keywords
- Auxiliary information
- Randomized response sampling
- Sensitive variable