Regularization through variable selection and conditional MLE with application to classification in high dimensions

It is often the case that high-dimensional data consist of only a few informative components. Standard statistical modeling and estimation in such a situation is prone to inaccuracies due to overfitting, unless regularization methods are practiced. In the context of classification, we propose a class of regularization methods through shrinkage estimators. The shrinkage is based on variable selection coupled with conditional maximum likelihood. Using Stein's unbiased estimator of the risk, we derive an estimator for the optimal shrinkage method within a certain class. A comparison of the optimal shrinkage methods in a classification context, with the optimal shrinkage method when estimating a mean vector under a squared loss, is given. The latter problem is extensively studied, but it seems that the results of those studies are not completely relevant for classification. We demonstrate and examine our method on simulated data and compare it to feature annealed independence rule and Fisher's rule.     

A ‘nondecimated’ lifting transform

Classical nondecimated wavelet transforms are attractive for many applications. When the data comes from complex or irregular designs, the use of second generation wavelets in nonparametric regression has proved superior to that of classical wavelets. However, the construction of a nondecimated second generation wavelet transform is not obvious. In this paper we propose a new ‘nondecimated’ lifting transform, based on the lifting algorithm which removes one coefficient at a time, and explore its behavior. Our approach also allows for embedding adaptivity in the transform, i.e. wavelet functions can be constructed such that their smoothness adjusts to the local properties of the signal. We address the problem of nonparametric regression and propose an (averaged) estimator obtained by using our nondecimated lifting technique teamed with empirical Bayes shrinkage. Simulations show that our proposed method has higher performance than competing techniques able to work on irregular data. Our construction also opens avenues for generating a ‘best’ representation, which we shall explore.     

The meaning of frailty according to Dutch older frail and non-frail persons

Frailty is a term often used by researchers and clinicians to describe a state in which older persons are at risk for adverse outcomes such as falls, disability, institutionalization and mortality. However, no study so far examined what frailty means to older persons. Therefore the aim of this study was to describe the meaning that older community-dwelling persons attach to frailty. Twenty-five semi-structured interviews were conducted. The interviews were analyzed using the grounded theory method. Frailty was described as being in poor health, having walking difficulties, feeling down, being anxious, having few social contacts and not being able to do the things one likes to do. Men described in more detail the physical dimension whereas women elaborated in more depth on the psychological and social component. Existing definitions of frailty should be adjusted to better reflect the meaning of frailty for older persons.     

On the margins of the child protection system: creating space for relational social work practice

In the UK, a threshold divides between two categories of children, child protection (CP) and child in need. Each category tends to be treated as a homogeneous entity, despite containing heterogeneous levels and forms of risk and need. CP practice, accompanied by regulation, protocols and procedures, aspires to achieve a coordinated multi‐agency response to identified concerns with available resources targeted towards this category. However, it is well known that those children assessed as falling just below the CP threshold can still have high levels of need and risk, requiring a level of social work involvement beyond the low‐resource and low‐oversight model that generally accompanies a child in need categorisation. This paper probes an approach to practice, which divides levels of risk within the child in need category enabling adequate, coordinated support and oversight to be provided for children and families with complex needs. Evidence from our study evaluating this approach suggests that a simple protocol provided a clear process within, which social workers and agency partners felt confident and safe to practice outside of the formal CP framework. The protocol prevented drift and helped to create a space within, which relational social work practice flourished.     

If You Want People to Listen to You,Tell a Story

We hypothesized that (a) when people share a meaningful story, as opposed to when they share information, they make their partner listen well, and (b) that narrative-induced listening is positively associated with speakers’ psychological safety and negatively associated with their social anxiety. In Study 1 (N = 45), we showed that a meaningful story is perceived much more as a narrative and higher in narrative quality than two types of informational-discourses (telling about daily routine and describing buildings). In Study 2 (N = 52), we randomly asked participants to either share a meaningful story or tell about their daily routine. The participants sharing a meaningful story reported that their interlocutor was a better listener, d = 0.61, 95% CI |0.32, 0.92|. In Study 3 (N = 42), we compared the effect of sharing a meaningful story to describing buildings, and replicated the results of Study 2, d = 1.10, 95% CI |0.61, 1.59|. Moreover, we found that the perceived listening, which was induced by the narrative, mediated the manipulation effects on psychological safety, and social anxiety. Thus, we concluded that when speakers share meaningful stories they make their partner listen well and consequently experience higher psychological safety and lower feelings of social anxiety.     

Male gamblers – suicide victims and living controls: comparison of adversity over the life course

The objective of this research is to investigate the possible difference in accumulation of adversity over the life course among decedent gamblers and living control gamblers in the province of Quebec, Canada. A sample of 90 adult men, 41 living participants and 49 suicide decedents, met the criteria for at-risk and problem gambling over the last five years. Data were obtained from informants in the case of suicide and directly from living controls during semi-structured face-to-face interviews using SCID I and II, SOGS, Module K and a recount of life trajectories methodology. The results showed a high level of adversity throughout the life trajectories of both groups. However, the living gamblers faced more difficulties during their lifespan than the suicide victims. In the last six months before their death, suicide victims had a significantly higher number of mental health disorders, especially co-morbid mental health disorders and anxiety, compared to the living controls. We also observed a rapid escalation of difficulties, more life events and more psychological distress in the last six months of their lives among the suicide decedents. Suicidal behaviour needs to be carefully investigated by medical staff and mental health practitioners, especially when gamblers demonstrate evidence of mental health disorders or financial difficulties.