On regression analysis using modified bivariate ranked set sampling design (BVRSS)

Abstract

Research on several forms of ranked set samples had been done by many researchers recently for estimating the population mean and other parameters. The results have ascertained that the ranked set samples are proven to be more efficient than the simple random samples in many situations. In this paper, modified bivariate ranked set sample (MBVRSS) was introduced to estimate a simple linear regression model when the regressors were assumed to be random. A primary data was collected on age and height in meter of 200 people from Ijebu-North Local Government Area of Ogun State. It is then shown that MBVRSS gives unbiased and highly efficient estimators of the regression model parameters, than those obtained by using bivariate simple random sampling (BVSRS), RSS or BVRSS using the same number of sample size. Key words: Modified bivariate ranked set sampling (MBVRSS), bivariate ranked-set sampling (BVRSS), bivariate simple random sampling (BVSRS) ranked set sampling (RSS), simple random sampling (SRS), simple linear regression.