Fast parallel $$\alpha $$-stable distribution function evaluation and parameter estimation using OpenCL in GPGPUs

\(\alpha \)-Stable distributions are a family of probability distributions found to be suitable to model many complex processes and phenomena in several research fields, such as medicine, physics, finance and networking, among others. However, the lack of closed expressions makes their evaluation analytically intractable, and alternative approaches are computationally expensive. Existing numerical programs are not fast enough for certain applications and do not make use of the parallel power of general purpose graphic processing units. In this paper, we develop novel parallel algorithms for the probability density function and cumulative distribution function—including a parallel Gauss–Kronrod quadrature—, quantile function, random number generator and maximum likelihood estimation of \(\alpha \)-stable distributions using OpenCL, achieving significant speedups and precision in all cases. Thanks to the use of OpenCL, we also evaluate the results of our library with different GPU architectures.     

Forward irrelevance

We investigate how to combine marginal assessments about the values that random variables assume separately into a model for the values that they assume jointly, when (i) these marginal assessments are modelled by means of coherent lower previsions and (ii) we have the additional assumption that the random variables are forward epistemically irrelevant to each other. We consider and provide arguments for two possible combinations, namely the forward irrelevant natural extension and the forward irrelevant product, and we study the relationships between them. Our treatment also uncovers an interesting connection between the behavioural theory of coherent lower previsions, and Shafer and Vovk's game-theoretic approach to probability theory.     

Transformed generalized linear models

The estimation of data transformation is very useful to yield response variables satisfying closely a normal linear model. Generalized linear models enable the fitting of models to a wide range of data types. These models are based on exponential dispersion models. We propose a new class of transformed generalized linear models to extend the Box and Cox models and the generalized linear models. We use the generalized linear model framework to fit these models and discuss maximum likelihood estimation and inference. We give a simple formula to estimate the parameter that index the transformation of the response variable for a subclass of models. We also give a simple formula to estimate the r$r$th moment of the original dependent variable. We explore the possibility of using these models to time series data to extend the generalized autoregressive moving average models discussed by Benjamin et al. [Generalized autoregressive moving average models. J. Amer. Statist. Assoc. 98, 214–223]. The usefulness of these models is illustrated in a simulation study and in applications to three real data sets.     

The effects of under‐skilling on need for recovery,losing employment and retirement intentions among older office workers: A prospective cohort study

Based on a sample of older workers from the Maastricht Cohort Study, the authors investigate the prevalence and dynamics of self‐reported under‐skilling and its effects on workers' need for recovery, their risk of losing employment and their retirement intentions over a two‐year period. They find that being under‐skilled is associated both with higher levels of need for recovery over time and with the risk of losing employment, but not with early retirement intentions. To achieve sustainable employment, they argue, the course of under‐skilling should be monitored throughout workers' careers, enabling timely interventions to avoid its negative consequences.     

Collecting feedback as a tool to reduce care paralysis: something for family group conferencing coordinators?

Family group conferencing (FGC) coordinators in public mental healthcare are confronted with clients who have little faith in professionals and organizations, who hold off decisions in their family life, who avoid care and who sometimes behave in a hostile manner. A lack of initiative to deal with their situation is not only reserved for clients – all bystanders, including professionals, can suffer from it. The multiplicity and severity of the client's problems lead to a situation wherein everyone involved waits for the initiative of the other. The independence of the FGC coordinator – a fellow citizen, who is free of organizational loyalties and comes to assist other citizens in establishing a plan – seems to work well with the client group of the public mental healthcare. However, the coordinator cannot always prevent deferral or failure of conferences. Drawing on empirical and theoretical findings, this paper considers the possibility of collecting feedback as a way to contribute positively to the alliance between FGC coordinators and those for whom a conference is deployed. We highlight findings from three case studies that centred on multiproblem families. The findings indicate the importance of feedback theory for FGC coordinators in enhancing trust and engagement.     

Racial/ethnic and socio‐economic biases in child maltreatment severity assessment in Spanish child protection services caseworkers

Evidence from several countries has shown the over‐representation of racial/ethnic minority groups in child protection services (CPS). The objective of the present study was to explore whether racial/ethnic and socio‐economic biases influence Spanish CPS caseworkers' judgements of the severity of child maltreatment. Moreover, the study attempted to explore the influence on these judgements of the use of structured instruments and professional experience. Two case vignettes of child maltreatment were presented to 405 CPS caseworkers and 169 students of social work and psychology. Family ethnic origin and income were manipulated in the vignettes. The findings showed no statistical evidence of biases related to family ethnic origin or socio‐economic status (SES) in Spanish CPS caseworkers' judgements of maltreatment severity. Biases related to family SES were found for students for the vignettes of physical abuse. CPS caseworkers and students who did not use a structured instrument to assess maltreatment severity tended to underestimate the severity for the vignettes of parental incapacity to control child/adolescent behaviour and to overestimate it for the vignettes of physical abuse. CPS caseworkers who used a structured instrument showed higher percentages of accuracy and inter‐rater agreement, supporting the relevance of structured tools in reducing potential caseworkers' biases.     

Longitudinal Factor Analysis of Public Expenditure Composition and Human Development in Brazil After the 1988 Constitution

The way public policies are managed in Brazil has changed since the 1988 Federal Constitution. This study aimed to identify how changes in the structure of public expenditure composition at Brazilian federal states influenced local human development in these states. The states’ public expenditures were categorized according to their nature as spending indices whereas human development was measured through a human development index (HD). To verify the relationships between these variables, an accounting-social theoretical model was created and estimated through latent growth modeling (LGM). The LGM measurement period comprised five administration cycles of the Brazilian states (1988–2011). Variables were measured on data of the second year of state government term; their mean initial values and growth rates were recorded. Results show that, influenced by policies of centralized regulation promoted by the federal government, only the social spending growth rates had statistically significant effect on the human development growth rate, although not considered of great magnitude. Among mean initial values, the most significant was that of minimum spending (SIm), which denotes the direct impact of the Fiscal Accountability Law [Lei de Responsabilidade Fiscal, in Portuguese] on human development improvements. The mean initial value of economic spending also showed a positive and significant effect on HD growth rate.     

Going solo: Findings from a survey of women aging without a partner and who do not have children

Greater longevity in the UK population has led to the increasing diversity of women experiencing aging in a multitude of ways. Internationally, gender inequalities in aging are still relatively invisible within both government policy and everyday life for particular groups of women. This article explores the concept of women growing older “solo”—by which we mean women who find themselves nonpartnered and aging without children as they move into later life. We report on the findings from a mixed-methods survey of 76 solo women in the UK aged 50 years and over, used to provide a broader overview of the issues and challenges they face as they move into later life. Qualitative data from the survey captured respondents’ perspectives about the links between their relationships status and well-being in later life and highlighted specific cumulative disadvantages emerging for some women as a result of their solo lifestyles. We discuss two key themes that were identified, “solo-loneliness” and “meaningful futures,” in conjunction with the relevant literature and make suggestions for future research within gender and aging studies that could enhance more positive approaches to solo lifestyles.     

Adjustment for nonresponse with variables from different sources: Bias correction and precision loss,with application to the Swiss European Social Survey 2012

ABSTRACT

Adjustment for nonresponse should reduce the nonresponse bias without decreasing the precision of the estimates. Adjustment for nonresponses are commonly based on socio-demographic variables, although these variables may be poorly correlated with response propensities and with variables of interest. Such variables nevertheless have the advantage of being available for all sample units, whether or not they are participating in the survey. Alternatively, adjustment for nonresponse can be obtained from a follow-up survey aimed at sample units which did not participate in the survey and from which the variables are designed to be correlated with response propensities. However, information collected through these follow-up surveys is not available for people in the sample who participated neither in the survey nor in its nonresponse follow-up. These two sets of variables when used in a nonresponse model for the Swiss European Social Survey 2012 differ only slightly with regard to their effect on bias correction and on the precision of estimates. The variables from the follow-up are performing slightly better. In both cases, the adjustment for nonresponse performs poorly.