Invariance In the linear model – an argument for χ2 and F in nonnormal situations

For a suitable norm, conservation of the distance between expectation and hypothesis may furnish a basis for data reduction by invariance in the linear, not neces-sarily normal, model. If the norm is Euclidean (i.e. based on some inner product), the maximal invariant is a pair of sums of squares. This provides support for traditional χ² (or F) - methods also in nonnormal cases. If the norm is l^p p≠2, or the supnorm, the maximal invariant is, at the best a air of order statistics.     

Approximate Bayesian Computation for Exponential Random Graph Models for Large Social Networks

We consider the issue of sampling from the posterior distribution of exponential random graph (ERG) models and other statistical models with intractable normalizing constants. Existing methods based on exact sampling are either infeasible or require very long computing time. We study a class of approximate Markov chain Monte Carlo (MCMC) sampling schemes that deal with this issue. We also develop a new Metropolis–Hastings kernel to sample sparse large networks from ERG models. We illustrate the proposed methods on several examples.     

Asymmetries in Information Processing in a Decision Theory Framework

Research in psychology suggests that some individuals are more sensitive to positive than to negative information while others are more sensitive to negative rather than positive information. I take these cognitive positive–negative asymmetries in information processing to a Bayesian decision-theory model and explore its consequences in terms of decisions and payoffs. I show that in monotone decision problems economic agents with more positive-responsive information structures are always better off, ex ante, when they face problems where payoffs are relatively more sensitive to the action chosen when the state of nature is favorable.      

Unwanted sexual experiences and sexual risks in gay and bisexual men: Associations among revictimization,substance use,and psychiatric symptoms

The data set of 2,182 gay/bisexual men's month‐long sexual diaries collected by the United Kingdom Socio‐sexual Investigations of Gay Men and Aids (Project SIGMA) was used to analyse the extent to which acts of anal intercourse (AI) are distributed among gay men, using the Lorenz/Gini methodology for measuring concentration. Most individuals (60%) who engage in AI do so only once or twice a month, but there is a long tail of those who do it much more. In terms of the amount of AI acts, one tenth of the individuals are performing half of the acts of AI. The Gini coefficient of concentration is high (0.55). Factors most affecting rates and concentration of risk behaviour include relationship status, HIV sero‐status and concordant/discordant partner status.     

Book review

W.G.Cochran: Sampling Techniques, 3rd. Ed. John Wiley & Sons, New York-Santa Barbara-London-Sydney-Toronto 1977. 428 S., £ 12.50; $21.50.

H.Toutenburg: Vorhersage in linearen Modellen. Akademie-Verlag, Berlin 1975, VIII, 176b S., 3 Tab., 28,– M.

O.Kallenberg: Random Measure. Akademie-Verlag, Berlin 1975; Academic Press, London 1976. 104 pp., 28,– M.     

Testing non-oblique hypotheses

This paper considers a likelihood ratio test for testing hypotheses defined by non-oblique closed convex cones, satisfying the so called iteration projection property, in a set of k normal means. We obtain the critical values of the test using the Chi-Bar-Squared distribution. The obtuse cones are introduced as a particular class of cones which are non-oblique with every one of their faces. Examples with the simple tree order cone and the total order cone are given to illustrate the results.     

Experimental designs and their efficiencies for spatially correlated observations in two dimensions

Optimality of experimental designs for spatially correlated observations is investigated.come two dimensional correlation structures are discussed and an attempt has been made to find optimal or nearly optimal design for each sitution.The solution lend to designs similar to that used for repeated measurements.The relative efficiency of the proposed designs in comparison to randomized latin square designs is tabulated for some cases.     

Rainfall variations and child mortality in the Sahel: results from a comparative event history analysis in Burkina Faso and Mali

Over the last 40 years, the Sahel has seen a long-term downward trend in rainfall. The importance of ecological variables as factors affecting child survival in rural subsistence societies has already been emphasized, but little empirical evidence has been gathered to support this. This paper presents a comparative event history analysis aimed at understanding how rainfall variations may influence child mortality in two neighbouring countries, Burkina Faso and Mali. These countries are similar in terms of population dynamics, economy, livelihood, child mortality and rainfall conditions (i.e. strong south–north decreasing rainfall gradient). Individual data for both countries came from two detailed nationally representative retrospective surveys conducted in 2000. Rainfall data for the 1960–1998 time period were obtained from the Climatic Research Unit. This study shows that child survival in each country is related to specific patterns of rainfall variation across livelihood regions, highlighting the complex nature of environmental causality of child mortality.     

Denn sie wissen was sie tun!

Information and communication technologies (IT) permeate both the business and the private world. They cause profound changes for companies in terms of structure, leadership, culture, education and the position of individuals within the context of the company. Established structures, decision making habits and methods of communication are being called into question. This means that communication is neither explicitly controllable nor in some cases even specified, but it also should not be viewed as arbitrary. Structures which are treated as contingent require meta-management, reflection and determination. Leaders are forced to act within a constellation of tensions between the reduction and the expansion of complexity as basic functions of structuring. Ideally this type of steering takes place consciously and with a view toward the future, taking into account the increasing complexity and dynamics of the environment and the corporate culture. Empirical studies of our four intervention research projects show that the situation for many companies is sobering. Companies cannot always entirely interpret the rapid development of communication technologies and the consequences it brings with it, nor can they completely make use of its potential advantages. Companies and individuals are forced to think actively within the networked economy. The demand for further education shifts between the company (expertise, employee loyalty) on the one hand and the responsibility of the individual for his or her own development on the other. Agreements are the key to bringing harmony and new solutions to the conflicts among the differing interests of individuals, subsystems and the company as a whole. Considering the complex requirements for companies and management caused by IT, this article shows that reflective observation of intervention research, guided by theory and with a focus on targeted change of management practice, is a process which provides new opportunities for action in both theory and practice. In this process, focusing (systemization) and reflection (enhancement of perspectives) are fundamental keys for dealing with complexity.     

Construction of New Three-Level Response Surface Designs

Response surface methodology is widely used for developing, improving, and optimizing processes in various fields. In this article, we present a method for constructing three-level designs in order to explore and optimize response surfaces combining orthogonal arrays and covering arrays in a particular manner. The produced designs achieve the properties of rotatability, predictive performance and efficiency for the estimation of a second-order model.