3 YR estimator
Yong You, J.N.K. Rao and Mike Hidiroglou
Previous | Next
WFQ applied the You and Rao (2002) method to model (2.1)
and obtained an estimator of given by
(3.1)
where is obtained from
(3.2)
by replacing by Note that the YR estimator (3.1) has the same
form as the EBLUP but it uses a non-optimal estimator for The YR estimators are self-benchmarking, i.e., since by (3.2)
However, the MSPE of will be slightly higher than the MSPE of based on because is asymptotically more efficient than
As in the case of the estimator of has and terms. We need to estimate the variance of in order to get the term in the estimator of the other terms and are not affected. It follows from (3.2) that
(3.3)
The estimator is obtained by substituting and for and in (3.3).
The estimator of is given by
(3.4)
where
The MSPE estimator (3.4) is nearly unbiased under
the true model (2.1), similar to the MSPE estimator (2.5) of
Remark:
Any estimator of may be adjusted as
for specified to satisfy the benchmarking constraint where In particular, we can use to obtain the adjusted EBLUP estimator. As
noted by WFQ, both and are estimators of the form because is equal to zero when is set equal to or Any set of estimators that satisfy has the self-benchmarking property.
Previous | Next