Selecting between two concomitant variables in ratio method of estimation

Abstract

The issues related to the choice of a concomitant variable when information is available on more than one concomitant variables, under the ratio method of estimation, are addressed. Several situations of preference are derived. The conditions thus obtained are tested on the wide variety of natural populations and the results indicate that the ratio estimator, based on the concomitant variable with the coefficient of variation closed as possible to the coefficient of variation of the study variable, may be preferred by efficiency arguments. © 2001 Elsevier Science Inc. All rights reserved.