scispace - formally typeset
Search or ask a question

Showing papers by "Rahul Mukerjee published in 1989"


Book
01 Jan 1989
TL;DR: In this paper, a factorial experiment involving n factors, F1, F2, F n, at m1, m2,..., m n (≥2) levels respectively, is considered and the effect due to this treatment combination is denoted by τ(j1,j2, j n ).
Abstract: Consider a factorial experiment involving n factors, F1, F2,..., F n , at m1, m2,..., m n (≥2) levels respectively. Let the levels of F i be coded as 0, 1,...,m i −1 (1 ≤ i ≤ n). A typical selection of levels j = (j1,j2,..., j n ), 0 ≤; j i ≤ m i , − 1, 1 ≤ i ≤ n, will be termed the jth treatment combination and the effect due to this treatment combination will be denoted by τ(j1,j2,..., j n ).

43 citations


Journal ArticleDOI
TL;DR: In this paper, the authors give methods for finding estimators with minimum model expected variance and the optimal strategy under a general correlated superpopulation model, and derive several earlier optimality results as special cases.
Abstract: Restricting attention to fixed size sampling designs and linear unbiased estimators of a finite population total, we give methods for finding estimators with minimum model expected variance and the optimal strategy under a general correlated superpopulation model. Some techniques popular in the theory of optimal experiments help in the derivation. Several earlier optimality results are deduced as special cases.

11 citations


Book ChapterDOI
01 Jan 1989
TL;DR: In this paper, a large class of tests, including the likelihood ratio, the conditional likelihood ratio (CRL), Rao's and Wald's tests, is considered. But no assumption has been made regarding curved exponentiality.
Abstract: We consider a very large class of tests containing in particular the likelihood ratio, the conditional likelihood ratio, Rao's and Wald's tests. In the presence of a nuisance parameter and under contiguous alternatives the third-order power of tests within the class has been studied. A third order optimality property of Rao's test has been proved. The performance of the conditional likelihood ratio test vis-a-vis the usual likelihood ratio test has been investigated. No assumption has been made regarding curved exponentiality.

9 citations


Book ChapterDOI
01 Jan 1989
TL;DR: In this article, it has been shown that under Stein's two-stage sampling procedure, the confidence interval for the population mean, as proposed by him, cannot be improved upon.
Abstract: It has been shown that under Stein's two-stage sampling procedure, the confidence interval for the population mean, as proposed by him cannot be improved upon. Allied results in the context of point estimation have also been indicated.

8 citations