Variance estimation using auxiliary information under a one unit per stratum sample design

Several variance estimators using auxiliary information were compared for estimating the variance of the total or ratio under a one unit per stratum sample design. The auxiliary information used consisted of data from the 1977 and 1982 Economic Census. One thousand probability proportional to size one primary sampling unit per stratum samples were selected for the Monte Carlo study of characteristics of interest in a content evaluation survey as part of the 1982 Economic Census program. Six variance estimators were applied to each sample and their bias and mean square errors were evaluated. The results are strikingly different between the variance estimators of the estimated total and ratio.     

Consistency of l 1 estimates in censored linear regression models

In this paper, the regression model with a nonnegativity constraint on the dependent variable is considered. Under weak conditions, L ₁ estimates of the regression coefficients are shown to be consistent.     

Estimation of the failure rate-a survey of nonparametric methods Part I: Non-Bayesian Methods

In this paper several nonparametric, non-Bayesian methods for estimating the failure rate function on which no monotonicity conditions have been imposed are surveyed. The survey attempts to consolidate and synthesize literature from several diverse areas of application, and endeavors to be as up-to-date as is feasible.     

Estimation of parameters in mixtures of inverse gaussian distributions

Iterative procedures are developed for finding maximum likelihood estimates of the parameters of mixtures of two inverse Gaussian distributions. The performance of the estimates based on small samples is studied by simulation experiments. Asymptotic efficiencies relative to estimates based on completely classified samples are also evaluated. Unless the components of the populations are widely separated, the maximum likelihood estimates perform poorly.     

Linear Models in a general parametric form

The linear model Y - N(Xb, σ²∑) with a restriction R'b = M'u + c is considered, where X, R, M, ∑ and c are known. Explicit formulae are obtained for the best linear unbiased estimator of K'b, for the F-test of the hypothesis K'b = W'v + a, and for the simultaneous confidence intervals of the parameters K′_i b' s, where K = [K₁,K₂,…K_s], w, and a are known, none of the matrices X, ∑, R, M, K, and W is required to have full ranks, and the design X can be one - or multi-way,complete or incomplete, balanced or not balanced, connected or disconnected.     

Peer Estimates of School-Aged Boys' and Girls'Aggression to Same- and Cross-Sex Targets

Research comparing boys' and girls' aggression has typically focused on the overall amount of aggression using a within-sex design. Less attention has been given to differences in children's aggression according to the sex of the target. In the research reported here, boys' and girls' amount and style of aggression were compared for same- and cross-sex targets. A peer estimation procedure was used with children in Grades 2, 6, 9, and 11. Physical, verbal, and indirect aggression were studied. Although there were some differences according to Grade level, both the amount and style of aggression for boys and for girls differed as a function of the sex of target, with cross-sex aggression generally falling between the amount and style of boys -to-boys and girls-to-girls aggression. The results draw attention to the possible separate nature of the cross-sex context, and highlight the importance of taking into account the target of aggression when investigating differences in aggression between boys and girls.     

Estimating P[X>Y] from exponential samples containing spurious observations

In this paper we discuss the problem of estimating P[X>Y] when X and Y are independent exponential random variables and the sample from each population contains one spurious observation. The estimates ate derived for exchangeable, identifiable and censored models and their performances are evaluated numerically.     

Regular mles for nonregular distributions

Distributions whose extremity values of the support depend on unknown pa¬rameters are usually known as nonregular distributions. In most cases, the MLEs for these parameters cannot be obtained by differentiation. Familiar examples are the uniform distribution on the interval (0,0) and the truncated exponential distribution with truncation parameter 0. However, there exist distributions whose extremity points of the support depend on unknown pa¬rameters, which nevertheless are regular in the sense that the MLEs can be obtained by differentiation. This note provides a method of constructing such nonregular distributions with regular MLEs.     

Recurrence relations for single and product moments of order statistics from a generalized half logistic distribution with applications to inference

In this paper, we derive several recurrence relations satisfied by the single and product moments of order statistics from a generalized half logistic distribution. These generalize the corresponding results for the half logistic distribution established by Balakrishnan (1985). The relations established in this paper will enable one to compute the single and product moments of all order statistics for all sample sizes in a simple recursive manner; this may be done for any choice of the shape parameter k. These moments can then be used to determine the best linear unbiased estimators of location and scale parameters from complete as well as Type-I1 censored samples.     

Bayes and minimax estimators for two-stage sampling from a finite population under measurement error models

We consider Bayes and Minimax estimates of population mean in sampling from a finite population in two-stages using simple random sampling without replacement at each stage, when the true values of the characteristic cannot be observed but only the values mixed with some measurement errors are observed. Minimax values of sample sizes have been found in case of equal-sized clusters and equal-sized second stage samples.