Investigation of an Empirical Probability Measure Based Test for Multivariate Normality


Book Description

Foutz (1980) derived a goodness of fit test for a hypothesis specifying a continuous, p-variate distribution. The test statistic is both distribution-free and independent of p. In adapting the Foutz test for multivariate normality, we consider using chi2 and rescaled beta variates in constructing statistically equivalent blocks. The Foutz test is compared to other multivariate normality tests developed by Hawkins (1981) and Malkovich and Afifi (1973). The set of alternative distributions tested include Pearson type II and type VII, Johnson translations, Plackett, and distributions arising from Khintchine's theorem. Univariate alternatives from the general class developed by Johnson et al. (1980) were also used. An empirical study confirms the independence of the test statistic on p even when parameters are estimated. In general, the Foutz test is less conservative under the null hypothesis but has poorer power under most alternatives than the other tests.







Testing For Normality


Book Description

Describes the selection, design, theory, and application of tests for normality. Covers robust estimation, test power, and univariate and multivariate normality. Contains tests ofr multivariate normality and coordinate-dependent and invariant approaches.







Proceedings of the Statistical Computing Section


Book Description

Papers presented at the annual meeting of the American Statistical Association ...




Energy Research Abstracts


Book Description

Semiannual, with semiannual and annual indexes. References to all scientific and technical literature coming from DOE, its laboratories, energy centers, and contractors. Includes all works deriving from DOE, other related government-sponsored information, and foreign nonnuclear information. Arranged under 39 categories, e.g., Biomedical sciences, basic studies; Biomedical sciences, applied studies; Health and safety; and Fusion energy. Entry gives bibliographical information and abstract. Corporate, author, subject, report number indexes.







Statistics in Industry


Book Description

This volume presents an exposition of topics in industrial statistics. It serves as a reference for researchers in industrial statistics/industrial engineering and a source of information for practicing statisticians/industrial engineers. A variety of topics in the areas of industrial process monitoring, industrial experimentation, industrial modelling and data analysis are covered and are authored by leading researchers or practitioners in the particular specialized topic. Targeting the audiences of researchers in academia as well as practitioners and consultants in industry, the book provides comprehensive accounts of the relevant topics. In addition, whenever applicable ample data analytic illustrations are provided with the help of real world data.




Advances on Theoretical and Methodological Aspects of Probability and Statistics


Book Description

At the International Indian Statistical Association Conference, held at McMaster University in Ontario, Canada, participants focused on advancements in theory and methodology of probability and statistics. This is one of two volumes containing invited papers from the meeting. The 32 chapters deal with different topics of interest, including stochastic processes and inference, distributions and characterizations, inference, Bayesian inference, selection methods, regression methods, and methods in health research. The text is ideal for applied mathematicians, statisticians, and researchers in the field.




Inside Chinese Organizations


Book Description

An empirical study was undertaken (a) to examine the characteristics, influencing factors, and effectiveness of a distinctive Chinese management model adopted by the newly founded private enterprises in mainland China, as well as (b) to test the "fit" notion in contingency theory. The study was based on a random sample of 124 standardized personal interviews with CEOs of Chinese industrial and service companies in Beijing and Shanghai. The standardized questionnaire mainly contained the Aston structure scales (Pugh and Hickson 1976) and Khandwalla's (1977) management schedules. Besides, three case examples were studied including one state-owned bank and two private enterprises. Using AMOS structural equation modeling, size, mass-technology and uncertainty, as well as CEO need for achievement, environmental constraints and selected Chinese characteristics, were specified as influencing factors of "bureaucracy" and managerial practices, which included strategic orientation, leadership style, decision-making, communication and subcontracting. Another model analyzed the influence of various key success factors on organizational effectiveness. The findings provided support for the re-adoption of a distinctive Chinese management model in private companies, characterized by autocratic leadership, low formalization, "Chinese entrepreneurship" and network-based "webs". Private enterprises also tended to readopt and emphasize traditional Chinese family-related values. The distinctiveness of this model was, ceteris paribus, due to this emphasis on familism, which had a direct impact on organizational design and an indirect influence through its negative association with company size, which in turn significantly covaried with organizational design. This family-based ownership model was positively associated with organizational effectiveness. As regards contingency theory, the "traditional" fit hypotheses could not be rejected. In addition, less effective companies were not necessarily in "misfit". "Appropriate" structure, together with an optimizing planning style and firm age, were positively associated with performance. Managerial recommendations were codified in a CHINA(c) strategy framework for high effectiveness.