Estimating Mean Of A Normal Population Using Known Coefficient Of Variation


B.BHASKARA RAMA SARMA*

* Faculty of Mathematics , BRS Classes, #2-284,Vivekananda street,Hanumannagar,Ramavarappadu,Vijayawada,A.P-521108; Mobile:9441924418;e-mail:bbramasarma@yahoo.co.in
 

ABSTRACT

            In estimating the parameters of a population if one can sacrifice the property of  unbiasedness, better estimators in view of minimum mean squared error(MMSE) can be obtained.Consider a normal population N ( ) with mean   and known coefficient of variation .Since the variance   is a function of  , it is appropriate if s is also considered along with the sample mean   for estimating  . Motivated by this observation, two estimators for   are proposed in this paper using   and s. The procedure adopted is based on that of Searle.D.T (1964). The large sample properties of these estimators and their comparison with that of the conventional estimator, i.e., sample mean  , are also investigated. The corresponding results are presented. Larger gains are observed in efficiencies for small sample sizes.

 

 


International eJournal of Mathematics and Engineering

Volume 1, Issue 4, Pages: 616 619