Bayesian Nonparametric Instrumental Variables Regression Based on Penalized Splines and Dirichlet Process Mixtures

We propose a Bayesian nonparametric instrumental variable approach under additive separability that allows us to correct for endogeneity bias in regression models where the covariate effects enter with unknown functional form. Bias correction relies on a simultaneous equations specification with flexible modeling of the joint error distribution implemented via a Dirichlet process mixture prior. Both the structural and instrumental variable equation are specified in terms of additive predictors comprising penalized splines for nonlinear effects of continuous covariates. Inference is fully Bayesian, employing efficient Markov chain Monte Carlo simulation techniques. The resulting posterior samples do not only provide us with point estimates, but allow us to construct simultaneous credible bands for the nonparametric effects, including data-driven smoothing parameter selection. In addition, improved robustness properties are achieved due to the flexible error distribution specification. Both these features are challenging in the classical framework, making the Bayesian one advantageous. In simulations, we investigate small sample properties and an investigation of the effect of class size on student performance in Israel provides an illustration of the proposed approach which is implemented in an R package bayesIV. Supplementary materials for this article are available online.     

Bayesian robustness of credible regions in the presence of nuisance parameters

The problem of finding the most robust γ-level credible region for the parameter of interest in the presence of a nuisance parameter, with respect to a class of ε-contaminated priors, is studied. The case of arbitrary con-taminations is first analyzed; it is proved that the most robust region for the parameter of interest is theγ-level highest marginal likelihood region (forγ ≥ 0.5). Then, the result is extended to any measurable (not necessarily one-to-one) function of the parameter. Finally, the case of contaminations assigning fixed probabilities to the sets of a partition of the parameter space is analyzed and a partial result is given.     

Sensitivity analysis in ordered and restricted parameter models

This paper deals with the problem of local sensitivity analysis in ordered parameter models. In addition to order restrictions, some constraints imposed on the parameters by the model and/or the data are considered. Measures for assessing how much a change in the data modifies the results and conclusions of a statistical analysis of these models are presented. The sensitivity measures are derived using recent results in mathematical programming. The estimation problem is formulated as a primal nonlinear programming problem, and the sensitivities of the parameter estimates as well as the objective function sensitivities with respect to data are obtained. They are very effective in revealing the influential observations in this type of models and in evaluating the changes due to changes in data values. The methods are illustrated by their application to a wide variety of examples of order-restricted models including ordered exponential family parameters, ordered multinomial parameters, ordered linear model parameters, ordered and data constrained parameters, and ordered functions of parameters.     

Latent groups and latent traits of participation in bullying via factor mixture analysis: An exploratory study

In this study, we drew on a person-centred approach to explore forms of involvement and roles in bullying. More specifically, we applied Factor Mixture Analysis (FMA) to an in-depth measure of bullying roles, with a view to identifying latent dimensions of participation in bullying and latent groups characterised by differential behavioural patterns. Participants were 899 high school students (51% females) attending the 9th (49%) and 11th grades (51%). The data analysis revealed six latent behavioural dispositions. Four of these overlapped with behavioural patterns typical of previously identified forms of participation in bullying, while two factors shed new light on bullying behaviours. We also identified four latent groups, whose profiles were only partially aligned with the roles previously theorised in the literature.     

Firm size and appropriability of the results of innovation

This article studies the influence of four dimensions of knowledge (codifiability, observability, complexity and dependency) on two types of appropriation methods: secrecy and other alternative methods (constant incremental improvements, lead time, moving down the learning curve and control of complementary resources) and how the firms’ size modifies these relationships.Based on a sample of 670 innovations developed by Spanish industrial firms, codifiability, observability and dependency were found to be the attributes that marks the effectiveness of both secrecy and the other alternative methods. Dependency marked the effectiveness of alternative methods in large companies and observability and codifiability in small firms.     

Clandestine existences and secret research: eliminating official discrimination in the Canadian military and going public in academia

The 1990s was a notable decade for lesbians and gays in the Canadian military. Two important changes were the 1992 elimination of the official policy permitting discrimination against homosexual service members, and the 1996 introduction of benefits to same-sex partners. These changes radically influenced the psychological day-to-day reality of lesbian and gay military members. Yet, given the military culture, lesbian and gay members only began to come out in significant numbers at the turn of the century. This article presents an overview of our experience with researching the history of lesbians and gays in the Canadian military during the late 1990s and early years of the new century. It reveals the early clandestine nature of our research, and recounts some of the trials, tribulations, resistance, and successes we encountered when dealing with ethics boards and funding sources. It also describes our relationship with the media and how the military actively took steps to stall our efforts. We draw parallels between our experiences as researchers, and those of the lesbian service-members we were interviewing.     

A national evaluation of Safe Schools/Healthy Students: outcomes and influences

The Safe Schools/Healthy Students (SS/HS) Initiative has awarded over $2 billion in grants to more than 350 school districts in partnership with local mental health, law enforcement, and juvenile justice agencies. To estimate the impact of grantee characteristics, grant operations, and near-term outcomes in reducing violence and substance use, promoting mental health, and enhancing school safety, logged odds ratios (LORs) were calculated contrasting Year 3 with Baseline performance from grantee-provided data on seven outcome measures. After comparing grantee performance across outcomes and outcomes across grantees, the LORs were entered as dependent variables in a series of meta-regressions in which grantee characteristics, grant operations, and near-term outcomes were tested after controlling for pre-grant characteristics. Findings indicate that the SS/HS Initiative significantly improved most outcomes, that within-grantee performance varied greatly by outcome, and that random-effects meta-regression appreciably decreased the variance available for modeling. The approach demonstrates that the SS/HS Initiative is effective and that locally collected performance data can be used to estimate grantee success in improving youth outcomes.     

Coping with obsessive relational intrusion and stalking: the role of social support and coping strategies

This study investigated the extent to which social support mediates negative effects of stalking and obsessive relational intrusion (ORI) victimization. A survey of 1,068 respondents indicated that (a) ORI/stalking victimization is positively related to negative symptoms and trauma; (b) five different types of coping responses are positively related to negative symptoms; (c) four domains of social support reveal small but significant negative relationships with negative symptoms; and (d) females are more threatened by unwanted pursuit than male victims, and male pursuers are more threatening than female pursuers. Structural equation modeling indicates that the influence of ORI/stalking on negative symptoms is mediated by the use of coping strategies and the adequacy of social support. Discussion speculates on the functional theoretical value of coping and support processes in managing unwanted pursuit and stalking.     

Towards a geography of transnational spaces: Indian transnational communities in Australia

Abstract Few studies on migrant transnationalism have explicitly adopted a geographical perspective, despite the widespread use of spatial metaphors in the literature and the potential advantages that a geographical approach offers. In this article, the geography of the transnational spaces of Punjabis, Kannadigas and Indo‐Fijians is analysed. Punjabis have constructed complex transnational spaces that are virtually global in scale. Kannadigas are engaged in transnational activities linking their places of residence with south India. Indo‐Fijians have emerged as a regional transnational community stretching across the Pacific Ocean. On the basis of their experiences, a consistent terminology is suggested and a typology of different models of transnational spaces is developed. This typology provides a tool to compare different transnational communities beyond the Indian experience. It can be seen as a preliminary step in the direction of a more theoretical approach that links the geography of migrant transnational spaces with sociological debates on social space.