Cases for the nugget in modeling computer experiments

Most surrogate models for computer experiments are interpolators, and the most common interpolator is a Gaussian process (GP) that deliberately omits a small-scale (measurement) error term called the nugget. The explanation is that computer experiments are, by definition, “deterministic”, and so there is no measurement error. We think this is too narrow a focus for a computer experiment and a statistically inefficient way to model them. We show that estimating a (non-zero) nugget can lead to surrogate models with better statistical properties, such as predictive accuracy and coverage, in a variety of common situations.     

THE EFFECTS OF CHANGING SAMPLE SIZE ON THE ALPHA AND BETA ERRORS: A PEDAGOGIC NOTE

This note discusses a theoretical peculiarity about the interrelationships between the sample size and the two types of decision errors in hypothesis testing theory. It is shown that for a given decision rule, while the alpha error invariably diminishes as the sample size becomes larger, the beta error could increase for certain values of the parameter under the alternative hypothesis.     

MATCHED PAIRS DISCRIMINATION: METHODOLOGY AND AN INVESTIGATION OF CORPORATE ACCOUNTING POLICIES

The matched-pairs methodology is becoming increasingly popular as a means of controlling extraneous factors in business research. This paper develops discriminant procedures for matched data and examines the properties of these methods. Data from a recent study by Hunt [14] on the determinants of inventory method choice are used to contrast the performance of the different methods. While all of the methods yield the same set of discriminating variables, those procedures that allow for the dependence among observations within a pair provide greater classificatory power than traditional multivariate techniques.     

One-mode projection analysis and design of covert affiliation networks

Within the field of national security and counterterrorism a great need exists to understand covert organizations. To better understand these cellular structures we model and analyze these cells as a collection of subsets of all participants in the covert organization, i.e., as hypergraphs or affiliation networks. Such a covert affiliation network structure is analyzed by evaluating the one-mode projection of the corresponding hypergraph. First we provide a characterization of the total distance in the one-mode projection using its corresponding cell-shrunken version. Secondly we evaluate the one-mode projection with respect to the secrecy versus information tradeoff dilemma every covert organization has to solve. We present and analyze affiliation networks representing common covert organizational forms: star, path and semi-complete hypergraphs. In addition we evaluate an example of a covert organization wishing to conduct an attack and compare its performance to that of the common covert organizational forms. Finally we investigate affiliation networks that are optimal in the sense of balancing secrecy and information. We show how any affiliation tree can be improved by altering its structure. Finally we prove that among covert organizational forms in the class of hypertrees with the same number of cells uniform star affiliation networks are optimal.     

The Effect of Attitudinal and Behavioral Commitment on the Internal Assessment of Organizational Effectiveness: A Multilevel Analysis

Using a multilevel analysis including 207 volunteers and paid workers nested within 51 nonprofit organizations (NPOs), this study examines the effect of individual and group attitudinal and behavioral commitment on their assessment of organizational effectiveness. Drawing on classical attitude theories, our results indicate that individuals with higher affective organizational commitment tend to assess their NPO’s effectiveness higher, while individuals staying because of the lack of alternatives assess it lower. However, in line with behavioral commitment theories, both relationships are mediated by the effect of teamwork behavioral commitment. We also found a negative effect of normative attitudinal commitment partially nested at the group level. Overall, our results suggest that encouraging volunteers and paid workers to participate in concrete teamwork behaviors on a daily basis constitutes a twofold benefit: it adds to the effect of affective attitudinal commitment at the individual level, while counter balancing the negative effects related to normative individual and collective resistances.     

Organizational Crisis Resistance: Examining Leadership Mental Models of Necessary Practices to Resist Crises and the Role of Organizational Context

This study explores which governance practices nonprofit leaders consider necessary to avoid organizational crises. Further, it explores whether these leadership mental models of crisis resistance depend on the organizational context. This helps determine whether practical learning points are organization specific or can be applied broadly. With a multilevel sample of 304 leaders from 44 Belgian nongovernmental development organizations, an exploratory path analysis reveals that nonprofit leaders consider continuous improvement, as a governance practice, particularly relevant for effective organizational crisis resistance. A multilevel analysis also shows that variations in leadership mental models cannot be explained by the organizational variables used in this study (organizational size, leadership group size, operational activities, and languages in the leadership group). This article concludes with a discussion of consequences for further research.     

An approximation to the distribution of the sample variance

A simple three-moment approximation is introduced for the distribution of the sample variance. Comparisons are given with other approximations discussed by Tan and Wong (1977) and with an approximation developed very recently by Mudholkar and Trivedi (1981).