Three-level meta-analysis of dependent effect sizes

Although dependence in effect sizes is ubiquitous, commonly used meta-analytic methods assume independent effect sizes. We describe and illustrate three-level extensions of a mixed effects meta-analytic model that accounts for various sources of dependence within and across studies, because multilevel extensions of meta-analytic models still are not well known. We also present a three-level model for the common case where, within studies, multiple effect sizes are calculated using the same sample. Whereas this approach is relatively simple and does not require imputing values for the unknown sampling covariances, it has hardly been used, and its performance has not been empirically investigated. Therefore, we set up a simulation study, showing that also in this situation, a three-level approach yields valid results: Estimates of the treatment effects and the corresponding standard errors are unbiased.     

Multicriteria Classification with Unknown Categories: A Clustering–Sorting Approach and an Application to Conflict Management

This work proposes an approach to cluster and sort a set of alternatives considering multi‐criteria categories with a partial order structure. It can be considered a heuristic approach because it does not attempt to derive an optimal partial order among all conceivable clusters of alternatives. Rather than this, it intends to be a simple approach that is transparent to the Decision Maker (DM) whose assistance is sought to help shaping the results. The approach proposed arises from the conjugation of traditional Clustering analysis and Multi‐criteria sorting tools. At the outset, the number of categories and their characteristics is unknown. First, we need to detect only the clusters themselves on the basis of a similarity measure independent of the preferences of the DM. Next, we detect potential partial order relations that might exist between them, according to the subjective preferences of the DM. Such preferences are elicited only after the DM has examined the clusters detected and deemed that these categories made sense. The new approach performs very well in a real‐world problem of management of intragroup conflicts and conflict handling strategies. Copyright © 2012 John Wiley & Sons, Ltd.     

Work Meaning Patterns in Early Career

Abstract

Work meaning patterns combine dimensions such as work centrality, expressive versus economic work goals, and entitlement versus obligation societal norms into a holistic picture of the evolution over time of the meaning of work. Data from a longitudinal study in eight countries are used to establish major work meaning patterns and to study their stability during early career paths. Further, some potential determinants of these work meaning patterns are analysed and their consequences for later careers are considered. Statistical analyses include: cluster analysis, multiple discriminant analysis, analysis of covariance combined with multiple classification analysis, analysis of variance, and chi-square analysis. Five cross-national work meaning patterns are identified for machine operators in their third year in the labour market. One-third of the sample had remained in the same work meaning pattern for 2 years, while two-thirds had changed their work pattern. Respondents' age, country, prior work environment, and prior work socialization behaviours and outcomes have an impact on work meanings held 2 years later. In addition the respondent's work meaning pattern allows subsequent career enhancing strategies and effort expenditure at the job to be predicted.     

Inattentional blindness for negative relationships in human causal learning

The present study focuses on the effect of selective attention on causal learning. Three effects of the level of attention to predictive symptoms in positive and negative contingency learning tasks are reported. First, participants accurately detected a positive relationship between an incidental cue and a contingent outcome, although judgements were slightly lower than those for the attended cue. Second, participants were unable to detect negative relationships between incidental cues and outcomes, which suggests a major role of selective attention in this type of learning. Third, participants retrieved the frequency of each trial type more accurately in the attended conditions than in the incidental conditions. These findings show how attention guides and constrains human causal learning and reveal an inattentional blindness effect for negative contingency learning.     

Subjective Situations and Logical Omniscience

The beliefs of the agents in a multi-agent system have been formally modelled in the last decades using doxastic logics. The possible worlds model and its associated Kripke semantics provide an intuitive semantics for these logics, but they commit us to model agents that are logically omniscient. We propose a way of avoiding this problem, using a new kind of entities called subjective situations. We define a new doxastic logic based on these entities and we show how the belief operators have some desirable properties, while avoiding logical omniscience. A comparison with two well-known proposals (Levesque's logic of explicit and implicit beliefs and Thijsse's hybrid sieve systems) is also provided.     

Psychiatric chrono‐epidemiology: its relevance for the study of aggression

Chrono‐epidemiology is the study of the distribution of health events according to time. The time intervals taken into account have cyclic patterns and include circadian, weekly, seasonal, and circannual rhythms. In the behavioural field, chrono‐epidemiology can be and has been applied to studies concerning the causes of death (suicide and homicide), clinical events such as admissions to and contacts with mental healthcare services, and the seasonal distribution of crime involving the use of violence. Frequently reported findings include clear circadian and seasonal rhythms present in the phenomenon of suicide (rhythms more evident in the case of violent suicide); seasonal rhythms in the expression of aggression by patients hospitalised for mental illness; and a circannual recurrence of crimes with a violent component. The intrinsic circadian and seasonal rhythms of some neuronal systems, particularly those of serotonin, involved in the control of mood and impulses, are thought to favour the behavioural rhythms observed, although the contribution of socio‐environmental factors, such as the fluctuation of supportive networks according to time, is also acknowledged. By demonstrating non‐casual recurrence of certain behaviour and by exploring the socio‐biological basis of the rhythms beneath these recurrences, chrono‐epidemiology may offer important etiologic and preventive clues to the understanding of the biological and environmental correlates of aggressive behaviour. Aggr. Behav. 28:477–490, 2002. © 2002 Wiley‐Liss, Inc.     

The Incompleteness of Theories of Games

We first state a few previously obtained results that lead to general undecidability and incompleteness theorems in axiomatized theories that range from the theory of finite sets to classical elementary analysis. Out of those results we prove several incompleteness theorems for axiomatic versions of the theory of noncooperative games with Nash equilibria; in particular, we show the existence of finite games whose equilibria cannot be proven to be computable.     

Exteroceptive control of response under delayed reinforcement

Three white rats, after 50 continuous reinforcements, were exposed successively, under dim illumination, to reinforcement delays of 1, 3, 5, 7.5, 10, 15, and 20 sec, with prolonged training at the 20-sec level. Behavior was maintained at each level, and an increase in interval was accompanied by an increase in post-reinforcement pause. Subsequently, under both 20- and 30-sec delays, the animals were tested during half of each daily session to determine the effect of introducing darkness during each delay interval. The result of this stimulus “support” was to regularize and increase response rate for each animal at both interval values.