|
Efficient class of hybrid estimators for population variance in two-phase sampling |
|---|---|
| รหัสดีโอไอ | |
| Creator | Prayas Sharma |
| Title | Efficient class of hybrid estimators for population variance in two-phase sampling |
| Contributor | Iqra Niaz, Aamir Sanaullah, Sat Gupta, Iram Saleem, Mohammad M. A. Almazah |
| Publisher | Maejo University |
| Publication Year | 2569 |
| Journal Title | Maejo International Journal of Science and Technology |
| Journal Vol. | 20 |
| Journal No. | 1 |
| Page no. | 91 |
| Keyword | auxiliary variable, two-phase sampling, ratio estimator, generalised hybrid estimator, variance estimation, mean square error |
| Website title | Maejo International Journal of Science and Technology |
| ISSN | 1905-7873 |
| Abstract | Auxiliary information is used to estimate population parameters and enhance the efficiency of estimators. However, when such information is partially or entirely unavailable, or when surveying a large sample is prohibitively expensive, two-phase sampling becomes practical. Unlike previous approaches, this study introduces a generalised hybrid estimator that integrates ratio, product and exponential-type strategies to achieve consistently lower mean squared errors. This unified framework extends existing variance estimation methods in two-phase sampling and demonstrates clear efficiency gains over conventional estimators. Therefore, this study aims at addressing the problem of variance estimation under such a context. This study suggests a class of efficient hybrid estimators for estimating finite population variance in two-phase sampling. The bias and mean squared error expressions are derived up to the first order of approximation and compared with those of existing estimators. A special case of the proposed class of estimators is also discussed. The performance of the proposed estimator is evaluated empirically using two real-world data sets. A simulation study is conducted using three bivariate normal populations with varying correlations between the auxiliary and study variables. To evaluate estimator performance, theoretical percentage relative efficiencies are computed. The results of both simulation and empirical studies demonstrate significant efficiency gains of the proposed estimator over existing approaches. |
