Some Compromised Exponential Ratio Type Imputation Methods in Simple Random Sampling
- PDF / 298,123 Bytes
- 13 Pages / 595.276 x 790.866 pts Page_size
- 86 Downloads / 162 Views
REVIEW ARTICLE
Some Compromised Exponential Ratio Type Imputation Methods in Simple Random Sampling Shakti Prasad1
Received: 10 October 2017 / Revised: 5 February 2020 / Accepted: 9 October 2020 Ó The National Academy of Sciences, India 2020
Abstract In this paper, some compromised exponential ratio type imputation methods have been suggested and their corresponding point estimators have been proposed in simple random sampling when some sample observations are missing. The expressions for the biases and mean square errors of the proposed point estimators have been derived. These estimators are compared with the mean imputation method, ratio imputation (Lee et al in J Off Stat 10:231–243, 1994) method, regression imputation method, compromised imputation (Singh and Horn in Metrika 51:267–276, 2000) method, Singh and Deo (Stat Pap 44:555–579, 2003) estimator, Singh (Statistics 43:499–511,
2009) estimator, Gira (Appl Math Sci 9(34):1663–1672, 2015) estimator, Singh et al. (Hacet J Math Stat 45(6):1865–1880, 2016) estimators and Adapted exponential ratio type (Bahl and Tuteja in J Inform Optim Sci 12(1):159–164, 1991) estimator. Simulation studies have been performed and the proposed estimators are better than some comparable estimators. Keywords Auxiliary variable Study variable Bias Mean square error (MSE) Percent relative efficiency (PRE) Non-response Mathematics Subject Classification 62D05
Significance Statement of Our Work
Many authors have considered the large number of different type of imputation methods and their corresponding estimators for non-response in survey sampling, but all such methods and their estimators may not be as precise as we desire, this required the need for new combinations which may produce desire results. In this paper, I considered compromised exponential ratio type imputation methods for all the missing values. Imputed values, using the proposed imputation methods, remain close to the true values in non-responding units. Our proposed methods provide better estimators of population mean using the known population values of auxiliary variable under SRSWOR and SRSWR Scheme. Therefore, it is recommended to use in future. & Shakti Prasad [email protected] 1
Department of Basic and Applied Science, National Institute of Technology, Arunachal Pradesh, Yupia, Papum Pare 791112, India
1 Introduction Non-response is one major problem, which is encountered by practitioners in the field of sample surveys. Determining the appropriate analytical approach in the presence of incomplete survey data due to non-response is a major question for the analysts and researchers, as the inference concerning population parameters can be spoiled if the suitable information about the nature of non-response is not known. A natural question arises what one needs to assume to justify ignoring the incomplete mechanism. Rubin [13] addressed three concepts: missing at random (MAR), observed at random (OAR) and parameter distribution (PD). Rubin defined ‘‘The data are MAR if the probability of the obs
Data Loading...
