Showing papers in "Calcutta Statistical Association Bulletin in 1982"
TL;DR: In this paper, a generalized estimator representing a class of estimators for the estimation of ratio and product of population parameters has been proposed, and a sub-class of optimum estimators from the proposed class has been investigated and it has been shown that every member of the subclass has the same minimum mean square error.
Abstract: A generalized estimator representing a class of estimators for the estimation of ratio and product of population parameters has been proposed. A subclass of optimum estimators from the proposed class has been investigated and it has been shown that every member of the subclass has the same minimum mean square error. Further the optimum value (depending upon population parameters) when replaced from sample values gives the estimators having the minimum mean square errors of the optimum estimators.
23 citations
TL;DR: It is proved that the universal optimality of fractional factorial plans derivable through orthogonal arrays in the presence of lower order interactions is proved.
Abstract: This paper extends the results of Cheng (1980) and proves the universal optimality of fractional factorial plans derivable through orthogonal arrays in the presence of lower order interactions. Also the connexion with orthogonal fractional factorial plans has been explored.
18 citations
TL;DR: Mukhopadhyay's (1974) sequential procedure for estimating the location parameter of a negnive exponential distribution is further analyzed in the case of point estimation as mentioned in this paper, and two main results are summarized in Theorems 1 and 2, and sbcse study asymptotic behaviors of the regret.
Abstract: Mukhopadhyay's (1974) sequential procedure for estimating the location parameter of a negnive exponential distribution is further analyzed in the case of point estimation. Two main results are summarized in Theorems 1 and 2, and sbcse study asymptotic behaviors of the “regret”.
16 citations
TL;DR: In this paper, the location parameter of an exponential distribution with known coefficient of variation is estimated using several estimators including the maximum likelihood estimator and the beacons estimator, including the Gaussian distribution.
Abstract: The plper considers estimation of the location parameter of an exponential distribution with known coefficient of variation. Several estimators including the maximum likelihood estimator and the be...
11 citations
TL;DR: In this paper, the problem of estimating the maximal (minimal) point of a quadratic response surface was considered, and the design was obtained by an heuristic approach motivated by the symmetry of the problem.
Abstract: The problem considered in this paper is that of estimating the maximal (minimal) point of a quadratic response surface. This problem was earlier considered by Chatterjee and Mandai (1981) where ‘deficiency of a design’ was used as a criterion for comparing different designs. Here we use the D-optimality crit rion in an extended sense and obtain designs under this criterion. Firstly, the design is obtained by an heuristic approach motivated by the symmetry of the problem. Finally, D-optimality is verified through a necessary and sufficient condition. The designs obtained are rotatable and are same as those for the deficiency criterion in a special case.
8 citations
TL;DR: In this article, a number of methods of construction of PBIB-RC designs, based upon the method of differences, are developed, which are used in the construction of the RRC.
Abstract: A number of methods of construction of PBIB-RC designs, based upon the method of differences, are developed.
6 citations
TL;DR: In this paper, the authors give an example of a curved exponential where the maximum likelihood estimate is not third order efficient either in the sense of Fisher-Rao or Rao, and show that it is not the case for any curve exponential.
Abstract: : The authors give an example of a curved exponential where the maximum likelihood estimate is not third order efficient either in the sense of Fisher-Rao or Rao.
4 citations
TL;DR: An estimator identical in form with the symmetrized Des Raj estimator based on a PPSWOR sampling design of size two was shown to be admissible within the class of all estimators of a finite population total for any fixed size sampling design as mentioned in this paper.
Abstract: An estimator identical in form with the symmetrized Des Raj estimator based on a PPSWOR sampling design of size two is shown to be admissible within the class of all estimators of a finite population total for any fixed size sampling design of size two.
2 citations
TL;DR: This paper has introduced some mixed order response surface designs like FORDSORD, FORDTORD, SORDTORD and discussed their analysis and construction and explained how they can be improved upon.
Abstract: In this paper we have introduced some mixed order response surface designs like FORDSORD, FORDTORD, SORDTORD and discussed their analysis and construction. We have also explained how we can improve upon a FORDSORD where p factors are first order and (v-p) factors are second order to get one in which p 1 factors are first order, p 2 factors are seeond order and other (v - p) factors are second order so that the design becomes a FORDGDSORD. We can then improve it further whenp, factors are first order and (v-p+p2) are second order.
2 citations
TL;DR: In this paper, a maximal distance statistic, based on the generalized pair chart, has been considered for testing the hypothesis of equality of the two distributions, and an alternative pair chart based on partial observations is also considered.
Abstract: The pair chart is known to be a convenient descriptive tool in comparing two samples and in calculating and interpreting various nonparametric procedures for the twosample problem (see Quade (1973)). To incorporate rightcensored data in the twosample problem, a generalized pair chart is considered and the same is utilized in the schematic computation of Gehan's Wstatistic (Gehan (1965)) and a triplet statistic (see Crouse and Steffens (1969)) among others. An alternative pair chart based on partial observations is also considered. A maximal distance statistic, based on the generalized pair chart, has been considered for testing the hypothesis of equality of the two distributions.
2 citations
TL;DR: In this article, the variance and the distribution of the Graybill-deal estimator of the common mean of two independent normal distributions with unknown variances are obtained in case of equal sample sizes.
Abstract: Sharp upper and lower bounds for the variance and the distribution of the Graybill-Deal estimator of ihe common mean μ of two independent normal distributions with unknown variances are obtained in case of equal sample sizes.
TL;DR: In this paper, the problem of construcHng fixed size sampling designs having constant inclusion probabilities of first two orders is considered, where the population units are subject to classification in one or more ways eliminating the chance of selection of samples for which the units have too lopsided a distribution over the classes.
Abstract: The problem considered in this paper is that of construcHng fixed size sampling designs having constant inclusion probabilities of first two orders. Some such sampling designs are developed in situations where the population units are subject to classification in one or more ways eliminating the chance of selection of samples for which the units have too lopsided a distribution over the classes.
TL;DR: The Rao-Rubin condition characterizing the Poisson distribution is further generalized to characterize certain Poisson-compound distributions as mentioned in this paper, which is a generalization of the condition characterised by the Gaussian distribution.
Abstract: The RaoRubin condition characterizing the Poisson distribution is further generalized to characterize certain Poissoncompound distributions.