摘要:

Ramey and Alam 1979 proposed a closed sequential procedure for selecting that one of k >2 multinomial events which has the largest probability. They provided tables of constants necessary to implement their procedure and, based on these constants, they calculated certain performance characteristics of their procedure. In the course of devising a fundamentally different procedure and studying its performance, we had occasion to check their constants and associated performance characteristics. We found quite a few of them to be incorrect. In the present paper we provide extended sets of correct constants and associated performance characteristics (some of which were not provided by Ramey and Alam), and describe the methods that we used to calculate them.

ISSN:0361-0918    

DOI:10.1080/03610918508812440

出版商:Marcel Dekker, Inc.    

年代:1985

数据来源: Taylor

摘要:

In this article we study the effect of truncation on the performance of an open vector-at-a-time sequential sampling procedure (P*B) proposed by Bechhofer, Kiefer and Sobel , for selecting the multinomial event which has the largest probability. The performance of the truncated version (P*BT) is compared to that of the original basic procedure (P*B). The performance characteristics studied include the probability of a correct selection, the expected number of vector-observations (n) to terminate sampling, and the variance of n. Both procedures guarantee the specified probability of a correct selection. Exact results and Monte Carlo sampling results are obtained. It is shown that P*BTis far superior to P*Bin terms of E{n} and Var{n}, particularly when the event probabilities are equal.The performance of P*BTis also compared to that of a closed vector-at-a-time sequential sampling procedure proposed for the same problem by Ramey and Alam; this procedure has here to fore been claimed to be the best one for this problem. It is shown that p*BTis superior to the Ramey-Alam procedure for most of the specifications of practical interest.

ISSN:0361-0918    

DOI:10.1080/03610918508812441

出版商:Marcel Dekker, Inc.    

年代:1985

数据来源: Taylor

摘要:

A Monte Carlo study was made of the effects of using simple linear regression, on the appropriate probability paper, to estimate parameters, quantiles and cumulative probability for several distributions. These distributions were the Normal, Weibull (shape parameters 1, 2, and 4) and the Type I largest extreme-value distributions. The specific objective was to observe differences arising from choice of plotting positions. Plotting positions used were i/(n+l), (i−3)/(n+.04), (i−.5)/n, either (i−.375)/(n+.25) or (i−.4)/(n+.2), and either F[E(Yi)] or F[E(£n Y)]. For each combination of 4 sample sizes (n=10(10)(40)), distribution, and plotting position, regression lines were found for each of N =9999 samples. Each regression line was used to estimate: (1) quantiles of 9 specific probabilities, (2) probabilities of 9 specific quantiles, and (3) return periods corresponding to 9 specific quantiles. Compa[rgrave]ison of the mean, variances, mean square error and medians of these estimates and of the regression coefficients confirm some results of Harter [Commun. Statist. A13(13), 1984] and provide further insight.

ISSN:0361-0918    

DOI:10.1080/03610918508812442

出版商:Marcel Dekker, Inc.    

年代:1985

数据来源: Taylor

摘要:

Two statistics are suggested for testing the equality of two normal percentiles where population means and variances are unknown. The first is based on the generalized likelihood ratio test (LRT), the second on Cochran's statistic used in the Behrens-Fisher problem. Size and power comparisons are made by using simulation and asympototic theory.

ISSN:0361-0918    

DOI:10.1080/03610918508812443

出版商:Marcel Dekker, Inc.    

年代:1985

数据来源: Taylor

摘要:

Maximum likelihood estimates and likelihood ratio statistics and their asymptotic null and nonnull distributions are derived for the k–population testing and estimation problem with patterned means and covariance matrices in the presence of missing data. These results are an extension of results of Szatrowski (1979) on the k–population complete data problem and Szatrowski (1983) on the one-population missing data problem. The standard delta method is the principal technique used for deriving the asymptotic nonnull distributions.

ISSN:0361-0918    

DOI:10.1080/03610918508812444

出版商:Marcel Dekker, Inc.    

年代:1985

数据来源: Taylor

摘要:

Previous simulations have reported second order missing data estimators to be superior to the more straightforward first order procedures such as mean value replacement. These simulations however were based on deterministic comparisonsbetween regression criteria even though simulated sampling is a random procedure. In this paper a simulation structured asan experimental design allows statistical testing of the various missing data estimators for the various regression criteria as well as different regression specifications. Our results indicate that although no missing data estimator is globally best many of the computationally simpler first order methods perform as well as the more expensive higher order estimators, contrary to some previous findings.

ISSN:0361-0918    

DOI:10.1080/03610918508812445

出版商:Marcel Dekker, Inc.    

年代:1985

数据来源: Taylor

摘要:

Approximations to the bootstrap estimate of bias and variance may be obtained by replacing the estimateto be bootstrapped by one which is linear, or,or quadratic,, in the resampling vector p. The bootstrap bias and variance ofandmay then be evaluated analytically. These estimators are discussed and then investigated via a Monte Carlo experiment whenis the least squares regression coefficient estimate. Included amongst these bias and variance estimates are the standard jackknife, Jaeckel's [1972] infinitesimal jackknife, and Hinkley's [1977] jackknife. Good performers included a quadratic infinitesimal jackknife and the Monte Carlo evaluated bootstrap when the number of replicates was about 6 or 7 times the sample size. Poor performers included the standard jackknife and Hinkley's. Further, when computational costs were held equal, bootstrapping by Monte Carlo did better than the jackknife.

ISSN:0361-0918    

DOI:10.1080/03610918508812446

出版商:Marcel Dekker, Inc.    

年代:1985

数据来源: Taylor

摘要:

An examination of the effects of inspection error on Bayesian predictive distributions is provided. Binomial and hypergeometric sampling models are considered, and predictive distributions for the number of observed defectives are derived. These predictive probabilities are tabulated for illustrative cases, and their impact on acceptance sampling is considered.

ISSN:0361-0918    

DOI:10.1080/03610918508812447

出版商:Marcel Dekker, Inc.    

年代:1985

数据来源: Taylor

摘要:

It has long been recognized that management is rewarded on the basis of achieving its goals. Thus, executive compensation is typically modeled in terms of sales, a proxy for firm growth, and profits, a proxy for firm financial performance. Histori-cally, economists have intensely debated whether executives are rewarded for maximizing profits (the classical hypothesis) or sales (the more recent corporate growth hypothesis). Empiri-cal support, using ordinary least squares (OLS) regression, has been offered for both positions with more recent studies suggesting that both positions may have merit. Unfortunately, the extreme multicollinearity that typically exists between profits and sales may render studies utilizing OLS regression inappropriate for resolving the debate. It is the purpose of this paper to contrast the results of OLS with two biased regression procedures (ridge regression and latent root regres-sion) in modeling the determinants of executive compensation.

ISSN:0361-0918    

DOI:10.1080/03610918508812448

出版商:Marcel Dekker, Inc.    

年代:1985

数据来源: Taylor

摘要:

The powers of several nonparametric tests for determining a treatment effect using matched censored data are compared by simulation. The tests include adaptations of two sample Prentice efficient score extensions of the Exponential Scores, Wilooxon and Normal Scores tests with three different variance estimators. The tests are compared on simulated exponential, loglogistic and lognormal matched censored data. No test performed uniformly better than the others. However the Exponential Scores, or Logrank, test using the permutation or hypergeometric variance seemed to be more powerful in quite a few situations as well as having other advantages.

ISSN:0361-0918    

DOI:10.1080/03610918508812449

出版商:Marcel Dekker, Inc.    

年代:1985

数据来源: Taylor