Advancing Treatment Strategies for Pathological Gambling

Experiences from treatment trials are described. The use of structured instruments for subject characterization and symptom change assessment is discussed and the challenges of treating specific groups (e.g., those with co-occurring alcohol use disorders) explored. Treatment for pathological gambling might be advanced most rapidly through collaborative multi-center efforts.     

Trust in Risky Messages: The Role of Prior Attitudes

Risk perception researchers have observed a "negativity bias" for hazard-related information. Messages indicating the presence of risk seem to be trusted more than messages indicating the absence of risk, and risk perceptions seem more affected by negative than positive information. Two experiments were conducted to examine alternative explanations of this finding within the area of food additives. Study 1 (N = 235) extended earlier work by (a) unconfounding message valence (positive or negative) from message extremity (definite or null finding) and (b) exploring the role of prior attitudes. Results suggested that negative/risky messages were indeed trusted more even when extremity was taken into account. However, prior attitudes significantly moderated the effect of message valence on trust. Positive messages were distrusted only by those with negative prior attitudes. Study 2 (N = 252), further explored the role of prior attitudes and extended the work by examining reactions to risky messages about a positively viewed additive--a vitamin. The results again found a moderating effect of prior attitudes on message valence. Participants had greater confidence in messages that were more congruent with their prior attitudes, irrespective of valence. Furthermore, positive messages had a greater impact on risk perception than negative messages. These findings suggest that greater trust in negative messages about hazards may be a product of a "confirmatory" rather than a "negativity" bias.     

On fair compensation

This paper analyses the problems arising in the pure exchange fair division model, when some dimensions of the resources are personal, fixed, and cannot be redistributed. The remaining resources must then be allocated in a compensatory way. A set of desirable normative properties is defined. No-envy satisfies these properties, but is not generally non-empty in this setting and other criteria are examined, for which existence results are given. General impossibility results obtain. In particular, it is generally impossible to compensate fully and only for differential personal resources, when preferences differ.     

Metropolitanizing Nonmetro Space: Population Redistribution and Emergent Metropolitan Areas, 1965–901

Abstract The recent volatility of population redistribution trends in the U.S. continues to stimulate the demographic imagination. This research sheds new light on this subject through an examination of newly designated metro areas, which collectively constitute the largest source of national metropolitan growth in two of the past three decades. The results show that, over this period, new metro areas have consistently drawn the majority of migrants from inter-state exchanges with existing metro territory and that these exchanges are greatest from census divisions that achieved metropolitan dominance prior to 1920. Employment data also reveal a strong positive association between FIRE activities and net in-migration. Together, these findings suggest that traditional processes of urban concentration, while present, now play a secondary role in the development of new metro centers, thus challenging conventional theories of metropolitanization where we might most expect them to apply.     

Decentralizing Risk Analysis in Large Engineered Systems: An Approach to Articulating Technical and Socioorganizational Dimensions of System Performance1

Nuclear power production in France, as the domain of one national utility and one reactor builder, has a standardized profile of technical installations, safety procedures, and personnel qualifications. Despite this relative homogeneity, discrepancies are observed from one production unit to another, notably in the area of worker-safety performance. There is a strong implication for risk analysis: varying performance cannot be attributed solely to the technical dimensions of equipment, procedures, and human skills. The authors retain as a working hypothesis that safety performance is an outcome of interactions between technical and organizational factors. Traditional risk analysis appears to be underequipped to represent such interactions. The notion of decentralized risk analysis (DRA) is introduced as a means of achieving this goal. A program of applied research carried out in a NPP facility is presented. It aims at increasing, across plant work structure, knowledge and control of these interactions specific to the given context. Systematic measurements and feedback of social representations are performed, using a three-dimensional factor space of individual and organizational values. Direct involvement of the total plant population, transfer of analytic tools, methodological continuity, interactive elaboration of data, coordination of different levels of findings with operations, follow-up and feedback of measured change, are the main features of this DRA process. In the case cited, improvement observed in safety performance parallels changes in measured social representations. The value of DRA for articulating technical and organizational dimensions of risk, and for integrating information into decision-making, is argued.     

Sur l'estimation du risque encouru en reconstruisant une forme aléatoire

The problem of estimating the risk corresponding to the reconstruction of a random pattern is reviewed. It is shown that for a particular but important model, the problem is reduced to the estimation of two parameters closely related to those appearing in a two-state Markov chain, which is of independent interest. The estimation of the Markov chain's parameters is studied from the decision-theoretic point of view. Estimators which are better than others previously considered are obtained and adapted to the estimation of the corresponding risk. Examples are are analyzed; even if a very empirical method is used to give values to the parameters of an a priorilaw, some good estimators of the risk are obtained.     

Social background,emotionality, and belief as predictors of punitiveness toward deviants

This research addresses the general question of which factors other than evidence determine the outcomes of criminal justice processing. Specifically, it investigates the relationships between social background factors, beliefs and emotions, and punitiveness toward deviants. The results offer support for the proposition that differing backgrounds are systematically linked with different views about human nature and society. The results also support the proposition that these variations in belief are linked to punitiveness toward deviants.