L’adaptation des modèles de production dans les pays émergents L’exemple de l’industrie automobile en Argentine

How to test for the existence of a new model of production based on the firm lending an ear to customer demands and valuing the individual and collective intelligence of its work force? The facts for putting this model to the test of reality come from an analysis of two Argentinean automobile-makers, subsidiaries of Toyota and Volkswagen. Various aspects of their organization policy provide evidence of their capacity for making adaptations by “hybridizing” the parent firm’s model with the rationales inherent in the network of local actors (suppliers and labor unions, in particular). These firms are open systems with a narrow margin of maneuver. — Special issue: Latin America.     

Street children in Trinidad and Tobago: understanding their lives and experiences

A community-based action research process, over a three-year period, employed interviews and participant observations with five children living on the streets in Port of Spain, the capital of Trinidad and Tobago. Data revealed violence in the family to be a contributing factor to children's presence on the streets: it was also a feature of their lives on the streets. A complex pattern emerged of violence, hostility, and distorted social interactions amidst negative attitudes and disdainful behaviour all around them. It is argued that, although it is not easy, more people – professionals and citizens – would do well to listen to, and understand the lives of, street children in order to be motivated for change.     

Heterogeneity and study size in random-effects meta-analysis

It is well known that heterogeneity between studies in a meta-analysis can be either caused by diversity, for example, variations in populations and interventions, or caused by bias, that is, variations in design quality and conduct of the studies. Heterogeneity that is due to bias is difficult to deal with. On the other hand, heterogeneity that is due to diversity is taken into account by a standard random-effects model. However, such a model generally assumes that heterogeneity does not vary according to study-level variables such as the size of the studies in the meta-analysis and the type of study design used. This paper develops models that allow for this type of variation in heterogeneity and discusses the properties of the resulting methods. The models are fitted using the maximum-likelihood method and by modifying the Paule–Mandel method. Furthermore, a real-world argument is given to support the assumption that the inter-study variance is inversely proportional to study size. Under this assumption, the corresponding random-effects method is shown to be connected with standard fixed-effect meta-analysis in a way that may well appeal to many clinicians. The models and methods that are proposed are applied to data from two large systematic reviews.     

Looking Forward in Candidate Gene Research: Concerns and Suggestions

Candidate Gene × Environment (cGxE) interaction research holds promise for helping us understand for whom and why environments matter for families and development. In their commentary on our target article (G. L. Schlomer, G. M. Fosco, H. H. Cleveland, D. J. Vandenbergh, & M. E. Feinberg, 2015), J. E. Salvatore and D. M. Dick (2015) present their view of the current state and future of cGxE research and frame the debate regarding its merits for advancing knowledge of gene–environment interplay. In this reply, we discuss points of agreement and departure and provide a list of 5 domains by which the quality of cGxE research should be evaluated. Our hope is that researchers will use this list as a guide for their own work.     

Non-parametric estimation of the long-range dependence exponent for Gaussian processes

We consider a class of long-range-dependent Gaussian processes defined in a semiparametric framework. We propose a new estimator of the long-range dependence parameter, based on the integration of the periodogram in two windows. We show that it is asymptotically Gaussian and calculate the rate of convergence. We optimise parameters defining the window function for the minimum mean-square-error criterion. In a Monte-Carlo study, we compare the proposed estimator with previously studied estimators.     

The Development of Students' Mathematics Self‐Concept in Relation to Gender: Different Countries,Different Trajectories?

Gender differences in the development of children's and adolescents' academic self‐perceptions have received increasing attention in recent years. This study extends previous research by examining the development of mathematics self‐concept across grades 7–12 in three cultural settings: Australia (Sydney; N=1,333), the United States (Michigan; N=2,443), and Germany (four federal states; N=4,688). Results of latent growth curve models document very similar patterns of self‐concept development in males and females in the three settings. First, gender differences in favor of boys were observed at the beginning of the observation period (grade 7). Second, gender was not significantly related to self‐concept change in either group, meaning that initial differences persisted across time. Third, the results provided no evidence that the form of the longitudinal change trajectories for mathematics self‐concept differed across the cultural settings. This pattern of results is inconsistent with explanatory models that predict converging or diverging gender differences in mathematics self‐concept. Furthermore, the results indicate that self‐concept development may be highly similar across western cultural settings.     

Free energy methods for Bayesian inference: efficient exploration of univariate Gaussian mixture posteriors

Because of their multimodality, mixture posterior distributions are difficult to sample with standard Markov chain Monte Carlo (MCMC) methods. We propose a strategy to enhance the sampling of MCMC in this context, using a biasing procedure which originates from computational Statistical Physics. The principle is first to choose a “reaction coordinate”, that is, a “direction” in which the target distribution is multimodal. In a second step, the marginal log-density of the reaction coordinate with respect to the posterior distribution is estimated; minus this quantity is called “free energy” in the computational Statistical Physics literature. To this end, we use adaptive biasing Markov chain algorithms which adapt their targeted invariant distribution on the fly, in order to overcome sampling barriers along the chosen reaction coordinate. Finally, we perform an importance sampling step in order to remove the bias and recover the true posterior. The efficiency factor of the importance sampling step can easily be estimated a priori once the bias is known, and appears to be rather large for the test cases we considered.