Search

Monte Carlo methods were used to study the performance of the largest-root test, two trace-type tests and three determinantal tests in the fixed-effects MANOVA model when certain assumptions are violated. Results indicated that for protection against nonnormality and heterogeneity of covariance matrices, the largest-root test should be avoided, while the Pillai-Bartlett trace test may be recommended as the most robust of the MANOVA tests, with adequate power to detect true differences in a variety of situations. [ABSTRACT FROM AUTHOR]

In this article we seek good estimators and designs for estimating a regression function, when the function is assumed to be a polynomial of given degree plus a "small" polynomial of given higher degree. We propose a procedure using designs and estimators which are linear combinations of standard designs and estimators, and verify that in the cases computed our procedure shows desirable robustness properties when compared with the standard procedure. [ABSTRACT FROM AUTHOR]

The robustness of T[sup 2] for samples of size 5, 10 and 20 from several bivariate distributions is investigated. Samples are presented from bivariate normal, uniform, exponential, gamma, lognormal and double exponential distributions. Related observations on the one sample t and paired t are also given. Highly skewed distributions resulted in too many extreme values of T[sup 2]. Other distributions gave conservative results. The use of the t-test and non-simultaneous techniques gave large overall levels of significance. [ABSTRACT FROM AUTHOR]

A general method of determining efficiency-robust estimation and testing methods is presented. This is used to derive admissible linear unbiased estimators (ALUEs) having the following optimality property: Over a specified (finite) family of possible symmetric shapes of the error distribution, the estimator's respective variances cannot be jointly improved. Such estimators are generalizations of Lloyd's best linear unbiased estimators (BLUEs) determined for single shapes. Simple illustrative examples are given. Other applications of the method to estimation and testing problems are discussed briefly. [ABSTRACT FROM AUTHOR]

In this paper the robustness of some of the existing tests for the exponential distribution when some outliers are present is reviewed. It is noted that the tests proposed by Epstein and Sobel [4] can be used without any major modification in the presence of outliers and a more suitable flexible test criterion is proposed for the exponential lower limit which is superior to Carlson's statistic [2] in all situations. The distribution of this statistic is derived and tables are constructed to facilitate the use of this statistic. Also some tests for outliers in the exponential case are proposed. [ABSTRACT FROM AUTHOR]

This research was funded, in part, by National Institutes of Health grant No. 5-T01-GM-01526. An earlier version of this paper was presented to a symposium, Methodology in Sociology, jointly sponsored by the Methodology Section of the American Sociological Association and Loyola University, Chicago, Illinois, June 10-11, 1970. The authors are especially grateful to G. William Walster for his helpful comments on the earlier version of the paper. [ABSTRACT FROM AUTHOR]

Long-range investment plans, particularly for expansion or decentralization, often consist of a number of separate but interacting investments spaced over a number of years. Where there is appreciable uncertainty about external conditions in the future, it is possible that a "best" scheme based on the state of current knowledge will prove during the course of the intervening years to be less than good. One way of guarding against this danger is to ensure that the early (and therefore irreversible) steps in the investment sequence keep open as many options of "good" schemes as possible. Concepts of robustness and stability are developed. As a practical example of the use of these concepts a case study is presented, which deals with the location of sites for new factories in an industrial expansion program. [ABSTRACT FROM AUTHOR]