Generalized Estimation Procedure in Two-Occasion Rotation Patterns
- PDF / 269,702 Bytes
- 9 Pages / 595.276 x 790.866 pts Page_size
- 60 Downloads / 312 Views
RESEARCH ARTICLE
Generalized Estimation Procedure in Two-Occasion Rotation Patterns G. N. Singh1 • Awadhesh K. Pandey1 • Mukti Khetan2
Received: 19 May 2018 / Revised: 29 May 2020 / Accepted: 14 August 2020 The National Academy of Sciences, India 2020
Abstract This paper describes some proficient estimation procedures in the presence of multi-auxiliary variables to enhance the precision of estimates in two-occasion rotation sampling. Utilizing information on several auxiliary variables, which are considered to be a positive correlation with the study variables on both occasions, we have suggested some better estimation procedures. The behaviors of the proposed estimation strategies have been examined along with the discussion of optimum replacement strategies, and the results obtained in these estimation procedures are demonstrated numerically through empirical and simulation studies, which present the supremacy in efficiencies of the proposed estimation procedures over the sample mean estimator, natural successive sampling estimator and recently developed estimator. Keywords Successive sampling Multi-auxiliary variables Mean squared error Optimum replacement strategy Simulation study Mathematics Subject Classification 62D05
& Awadhesh K. Pandey [email protected] G. N. Singh [email protected] Mukti Khetan [email protected] 1
Department of Mathematics and Computing, Indian Institute of Technology (Indian School of Mines), Dhanbad 826004, India
2
PG Department of Statistics, Sambalpur University, Sambalpur, Odisha 768019, India
1 Introduction The appropriate use of information on supplementary variable in probability sampling offers considerable improvement in the efficiencies of the estimators of population parameters such as population mean, population median and population variance. The estimation of finite population mean attracts the attention of survey practitioners for several practical applications in different fields of socioeconomic, agricultural and environmental sciences, where the various characters are liable to change over the period of time. For instance, a specialist or owner involved in the tobacco business might be intrigued (1) to know the average or aggregate deals amid various seasons, (2) to know the pattern of change in average or total sales over two different seasons or (3) to know both the issues simultaneously. For such situations, continuous monitoring is required because one-time surveys do not provide the required information. Rotation (successive) sampling is the appropriate method to deal with such problems where the samples are selected following a specific rule, with partial replacement of units on different occasions. To meet these requirements, rotation (successive) sampling is the most appropriate statistical tool to generate the reliable estimates of population parameters on different occasions. Jessen [1] introduced the procedure of using information available on the first occasion to enhance the precision of estimates. Later on, the theory of rotat
Data Loading...
