Abstract
In this paper, we discuss the estimators based on geometric and harmonic mean to estimate the population mean in a finite population based on simple random sampling design when information on several auxiliary characters negatively correlated with the main variable is available. Bias and MSB of these estimators have been obtained and it is shown that these are less biased than Singh's (1967) estimate based on arithmatic mean under usually satisfied practical situations.
Get full access to this article
View all access options for this article.