Robust likelihood functions in Bayesian inference

In order to deal with mild deviations from the assumed parametric model, we propose a procedure for accounting for model uncertainty in the Bayesian framework. In particular, in the derivation of posterior distributions, we discuss the use of robust pseudo-likelihoods, which offer the advantage of preventing the effects caused by model misspecifications, i.e. when the underlying distribution lies in a neighborhood of the assumed model. The influence functions of posterior summaries, such as the posterior mean, are investigated as well as the asymptotic properties of robust posterior distributions. Although the use of a pseudo-likelihood cannot be considered orthodox in the Bayesian perspective, it is shown that, also through some illustrative examples, how a robust pseudo-likelihood, with the same asymptotic properties of a genuine likelihood, can be useful in the inferential process in order to prevent the effects caused by model misspecifications.     

Simple non-parametric estimators for unemployment duration analysis

Summary.  We consider an extension of conventional univariate Kaplan–Meier-type estimators for the hazard rate and the survivor function to multivariate censored data with a censored random regressor. It is an Akritas-type estimator which adapts the non-parametric conditional hazard rate estimator of Beran to more typical data situations in applied analysis. We show with simulations that the estimator has nice finite sample properties and our implementation appears to be fast. As an application we estimate non-parametric conditional quantile functions with German administrative unemployment duration data.     

Research note: Problems in monitoring the safety performance of British manufacturing at the end of the twentieth century

I have frequently expressed the view that the statistics for over three day injuries cannot be relied upon for the sort of temporal research on safety in which I have recently been engaged, for instance concerning the effectiveness of the 1974 Health and Safety at Work Act (Nichols, 1989a; 1990). It has been suggested in the context of some related research (Nichols, 1989b; 1991a) that the over three day injury rate may actually be a useful indicator of safety in some respects and that the fatality rate may not be (Tombs, 1992). By contrast, the following short account looks to the future and considers whether, in the 1990s, a point has been reached when any official injury statistics for employees in UK manufacturing can provide satisfactory monitoring of improvements or otherwise in safety over time. First, the fatality rate is briefly discussed, then the rate for major injuries, then the rate for over three day injuries. Two other new sources of injury data are also briefly considered, the 1990 Labour Force Survey (LFS) and the 1990 Work Industrial Relations Survey (WIRS3).     

"Take Me Along"-A Marital Syndrome

The work-identity-marriage relationship for females and males is examined. Physicians and their wives illustrate some of the patterns of a "take me along" marital syndrome involving contingent identity for the females. Implications are discussed.     

Living on the margins: older homeless adults in Toronto

A handful of scholars have acknowledged that, along side the traditional homeless, there are now older people who become homeless for the first time in old age. Few researchers, however, have systematically compared the recent older homeless with the chronic or traditional homeless. In the research presented here, we compare recent older homeless with long-term older homeless adults in Toronto according to their health and wealth, their housing history, and their use of health and social services. Findings indicate that people who become homeless for the first time at older ages have needs that are different from the lifetime elderly homeless and require different approaches to intervention.     

Factors Related to Social Competence in Elementary School Among Children of Adolescent Mothers

This study examined factors related to social competence in first grade among children of low‐income, adolescent mothers. The sample included 83 mother–child dyads who had participated in a family support program for adolescent mothers. Characteristics of the child, mother, and the child's living context were examined for their relationship to children's social skills and problem behaviors. Several factors had significant bivariate relationships with social competence. Children with higher social skills and lower levels of problem behaviors were more likely to have received higher quality parenting, to have higher academic skills, and to live in neighborhoods with lower poverty rates. Children with higher social skills were less likely to have changed schools. In a path analysis, only academic skills were significantly related to either measure of social competence when other factors were controlled. Quality of parenting was indirectly related to social competence, and the effect was mediated by academic skills. School transience was also related to academic skills in the path analysis.