Assessing and Improving the Performance of Nearly Efficient Unit Root Tests in Small Samples

Abstract

The development of unit root tests continues unabated, with many recent contributions using techniques such as generalized least squares (GLS) detrending and recursive detrending to improve the power of the test. In this article, the relation between the seemingly disparate tests is demonstrated by algebraically nesting all of them as ratios of quadratic forms in normal variables. By doing so, and using the exact sampling distribution of the ratio, it is straightforward to compute, examine, and compare the test' critical values and power functions. It is shown that use of GLS detrending parameters other than those recommended in the literature can lead to substantial power improvements. The open and important question regarding the nature of the first observation is addressed. Tests with high power are proposed irrespective of the distribution of the initial observation, which should be of great use in practical applications.     

Bullying and Victimization Among Young Elementary School Children: The Role of Child Ethnicity and Ethnic School Composition

School-aged children with an ethnic minority background are relatively often involved in bullying and victimization, but the role of ethnic composition of schools in this context remains unclear. This study examined the relation between ethnic minority background, ethnic school composition, and bullying behaviour around primary school entry in the Netherlands. The study was based on a 2008/2009 school survey in Rotterdam, a Dutch city where about 50 % of children have a non-Dutch background. For 8523 children, teacher reports of bullying behaviour at age 5–6 years were available. Children with a non-Dutch background had higher odds of being a victim (adjusted OR 1.41, 95 % CI 1.11, 1.80), bully (OR 1.38, 95 % CI 1.20, 1.58) or bully-victim (OR 1.38, 95 % CI 1.19, 1.62) than children of Dutch national origin. Ethnic diversity in schools increased children’s risk of bullying behaviour (e.g. OR_victim per 0.1 increase in 0–1 diversity range = 1.06, 95 % CI 1.00, 1.13), with children of both Dutch and non-Dutch national origin relatively more often involved in bullying in ethnically diverse schools. The proportion of same-ethnic peers in school reduced the risk of bullying among children of Dutch national origin (e.g. OR_victim per 10 % more same-ethnic children = 0.90, 95 % CI 0.83, 0.98), but not among non-Dutch children. In conclusion, ethnic minority background and ethnic diversity within schools are risk factors for bullying among 5–6 year olds. Plausibly, reductions in absolute numbers of bullying events may be obtained with tailor-made interventions in ethnically diverse schools. Such interventions should preferably be offered early in the school curriculum.     

Poverty Dynamics among Recent Immigrants to Canada1

This paper examines two issues: (1) poverty dynamics among successive cohorts of entering immigrants to Canada, and (2) whether rising educational attainment and increasing share in the “skilled” class since the early 1990s has resulted in improvements in poverty entry, exit, and chronic poverty. The entry to poverty is very high during the first year in Canada, but low in subsequent years. The dramatic move toward more labor‐market friendly characteristics of entering immigrants had only a very small effect on poverty outcomes, in part because the relative advantage of holding a degree diminished, and “skilled economic” class immigrants were more likely to enter poverty than their “family” class counterparts.     

Credibility of PR practitioners: The impact of professional journalism background on trustworthiness,expertness, and homophily evaluations

This study demonstrates that a professional journalism background contributes to the credibility evaluations of public relations practitioners in regards to expertness and the capability to meet journalistic values and expectations. However, having a professional journalism background does not necessarily render public relations practitioners any more trustworthy than not having such credentials.     

Beyond normalization and impairment: theorizing subjectivity in learning difficulties – theory and practice

Normalization and social role valorization continue to play a central role in shaping debates and practice relating to learning difficulties. In the context of recent arguments this paper draws on the work of Foucault to deconstruct these theories. Foucault’s work alerts us to a conceptual confusion at their heart which reproduces a common but problematic individual–society dualism. There is an implicit, and problematic, presence in the theories of a pre‐social individual conceived as having essential impairments and who is passive in the face of negative socialization. We propose that Foucault’s ‘ethical’ domain of inquiry, with its concern for how people actively understand themselves and govern their conduct in relation to specific values and a ‘truth’ that they are obliged to recognize in themselves, provides the basis for returning the individual‐as‐subject to theories in an active, critical manner.     

Computationally simple estimation and improved efficiency for special cases of double truncation

Doubly truncated survival data arise when event times are observed only if they occur within subject specific intervals of times. Existing iterative estimation procedures for doubly truncated data are computationally intensive (Turnbull 38:290–295, 1976; Efron and Petrosian 94:824–825, 1999; Shen 62:835–853, 2010a). These procedures assume that the event time is independent of the truncation times, in the sample space that conforms to their requisite ordering. This type of independence is referred to as quasi-independence. In this paper we identify and consider two special cases of quasi-independence: complete quasi-independence and complete truncation dependence. For the case of complete quasi-independence, we derive the nonparametric maximum likelihood estimator in closed-form. For the case of complete truncation dependence, we derive a closed-form nonparametric estimator that requires some external information, and a semi-parametric maximum likelihood estimator that achieves improved efficiency relative to the standard nonparametric maximum likelihood estimator, in the absence of external information. We demonstrate the consistency and potentially improved efficiency of the estimators in simulation studies, and illustrate their use in application to studies of AIDS incubation and Parkinson’s disease age of onset.     

Probabilistic linear solvers: a unifying view

Statistics and Computing - Several recent works have developed a new, probabilistic interpretation for numerical algorithms solving linear systems in which the solution is inferred in a Bayesian...     

Pauvreté et convergence des consommations au Canada*

To study changes in household consumption patterns, five socio‐economic household groups were defined using a new multidimensional index (IMPR) composed of three dimensions: satisfaction of basic needs, marginalization relative to a reference population, and total disposable household income. When household incomes rose from 1969 to 1992, lower socio‐economic groups did not display the same consumption behaviours as upper‐income classes. The growth in household income during their life‐cycles did not allow them to catch up to higher‐income households. Middle‐class households experienced marked changes in consumption over their life‐cycle's, but in the 1990s, they experienced difficulties. Cinq groupes socioéconomiques de ménages sont définis à l'aide d'un nouvel indice multidimensionnel de pauvreté‐richesse (IMPR), construit à partir de trois dimensions: satisfaction des besoins de base, marginalisation par rapport à une population de référence et revenu total disponible du ménage. De 1969 à 1992, quand les revenus des ménages augmentent, les classes socioéconomiques qui occupent des positions inférieures n'ont pas les mêmes comporte‐ments de consommation que les classes supérieures. La croissance des revenus des ménages les moins riches au cours de leur cycle de vie ne leur permet pas de rattraper les positions des ménages plus riches. Dans les années 1990, la classe moyenne connaît une situation difficile.     

Suicide ideation among US college students. Associations with other injury risk behaviors

Suicide, the endpoint of a continuum that begins with suicide ideation, is the third leading cause of death among the US college-aged population. The first and second leading causes of death among this age group, unintentional injury and homicide, may also be linked to suicide ideation. We used data from the National College Health Risk Behavior Survey to examine the association between suicide ideation and injury-related behaviors among 18- to 24-year-old college students. Students who reported suicide ideation were significantly more likely than students who did not report considering suicide to carry a weapon, engage in a physical fight, boat or swim after drinking alcohol, ride with a driver who had been drinking alcohol, drive after drinking alcohol, and rarely or never used seat belts. Given this clustering of injury-related risk behaviors, college prevention programs should aim to reduce risks for injuries comprehensively, rather than addressing each risk behavior separately.