Remainder Systematic Markov Chain Design

Fei-Fei Kao ； Ching-Ho Leu ； Chien-Hao Ko

《International Journal of Information and Management Sciences》 22：4 (2011/12) Pp. 327-342

https://doi.org/10.6186/IJIMS.2011.22.4.1

並列摘要

Systematic sampling is one of the simplest, easiest and the most common used sampling methods. However, when the population size N is not a multiple of the sample size n, the systematic sampling cannot be performed. Not only is it difficult to determine the sampling interval k, but the sample mean will be a biased estimator of the population mean. To solve this problem, this paper proposes an improved method for the systematic sampling: the remainder systematic Markov chain design. The first- and second-order inclusion probabilities are derived, yielding the Horvitz-Thompson estimator and its variance. The simulation results demonstrate the effectiveness of the proposed method for different super-populations.

並列關鍵字

Systematic sampling ； Markov chain design ； remainder linear systematic sampling ； Horvitz-Thompson estimator

參考文獻

Breidt, F. J.(1995).Markov chain designs for one-per-stratum sampling.Survey Methodology.21,63-70.

Chang, H. J.,Huang, K. C.(2000).Remainder linear systematic sampling.Sankhyā B.62,249-256.

Cochran, W. G. (1946). Relative accuracy of systematic and stratified random samples for a certain class of populations, The Annals of Mathematical Statistics, Vol.17, pp.164-177.

Cochran, W. G.(1977).Sampling Techniques.New York:Wiley.

Das, A. C.(1950).Two-dimensional systematic sampling and the associated stratified and random sampling.Sankhya.10,95-108.