Adaptive ranked-set sampling with multiple concomitant variables: an effective way to observational economy

Abstract

The notion of ranked-set sampling (RSS) proposed by McIntyre provides an effective means of achieving observational economy in certain particular situations. The use of concomitant variables broadens the range of application of RSS. In this paper, we deal with RSS with multiple concomitant variables and develop an adaptive RSS procedure in a general context. The particular case of multiple linear regression estimates with RSS is treated in detail. The procedure is illustrated with a real data set. Some simulation results are also given.