1Statistics and Information Department, Minister's Secretariat, Ministry of Health, Labour and Welfare, Kasumigaseki, Chiyoda-ku, Tokyo, JAPAN
American Journal of Applied Mathematics and Statistics.
2014,
Vol. 2 No. 5, 344-351
DOI: 10.12691/ajams-2-5-8
Copyright © 2014 Science and Education PublishingCite this paper: Fumito Muguruma. Roulette Model of Systematic Sampling.
American Journal of Applied Mathematics and Statistics. 2014; 2(5):344-351. doi: 10.12691/ajams-2-5-8.
Correspondence to: Fumito Muguruma, Statistics and Information Department, Minister's Secretariat, Ministry of Health, Labour and Welfare, Kasumigaseki, Chiyoda-ku, Tokyo, JAPAN. Email:
muguruma-fumito@mhlw.go.jpAbstract
Considering a natural model of systematic sampling, we can calculate the expectation of statistic exactly as it is without regarding it as any other sampling methods. I will demonstrate the details of calculation and get some results. It can be shown that the sample mean of systematic sampling is an unbiased estimator of population mean. Variance of sample mean can be described explicitly and depends on how to sort a population list. Sum of the variance of sample mean and the mean of sample variance is kept constant between random and systematic sampling. As the sample size becomes larger, the variance of sample mean converges to 0.
Keywords