Estimation in Mail Surveys Under PSNR Sampling Scheme
by
D. Shukla
Department of Mathematics and Statistics, Dr. H.S. Gour University, Sagar, MP, India
Coauthors: Jayant Dubey (Department of Mathematics and Statistics, Dr. H.S. Gour University, Sagar, MP, India)

In the set up of a stratified sampling scheme, when frames of each stratum are not available completely due to unavoidable reasons then survey practitioners are being advised to use post stratification scheme for the estimation of population parameters. Suppose these post startified units have only a few respondents and a large number of non-respondents then estimation of population parameter faces a trouble. This is most likely situation in mail survyes when questionnaires are mailed to respondents and they are supposed to return these, after completion, earlier to a prefixed deadline. This paper presents a new scheme named "Poststratified non-response (PSNR)" sampling scheme to handle a high rate of non-response likely to occur in mail surveys. The estimation of the population mean is an object to proceed for and the PSNR scheme has several steps to complete before estimation. An unbaised estimator has been found existing and precision is examined. A usual approach of cost analysis is incorporated to estimate the optiaml sample size required under prefix cost. Due to certain limitations of this usual approach, a new alternative approach is proposed and observed a solution oriented for cost optimal sample size. All the derived results are numerically supported and the PSNR scheme is found effective and useful.

Date received: November 14, 2000

Copyright © 2000 by the author(s). The author(s) of this document and the organizers of the conference have granted their consent to include this abstract in Atlas Conferences Inc. Document # cafx-05.