On low-dimensional projections of high-dimensional distributions

Let P be a probability distribution on q -dimensional space. The so-called Diaconis-Freedman effect means that for a fixed dimension d<

Stochastic search for semiparametric linear regression models

This paper introduces and analyzes a stochastic search method for parameter estimation in linear regression models in the spirit of Beran and Millar [Ann. Statist. 15(3) (1987) 1131–1154]. The idea is to generate a random finite subset of a parameter space which will automatically contain points which are very close to an unknown true parameter. The motivation for this procedure comes from recent work of Dümbgen et al. [Ann. Statist. 39(2) (2011) 702–730] on regression models with log-concave error distributions.

Social Factors Associated with Readiness for Sexual Activity in Adolescents: A Population-Based Cohort Study

Various factors are associated with sexual activity in adolescence and it is important to identify those that promote healthy and adaptive romantic and sexual development. The objectives of this study were to describe rates of early sexual intercourse (before 16 years) and sexual readiness in adolescence and to assess the extent to which these were social patterned. We prospectively studied nearly 5,000 15-year-olds from the Avon Longitudinal Study of Parents and Children, a UK birth cohort. Between 2006 and 2008, female and male participants answered a computer assisted interview about romantic and sexual behaviors in the last year. Predictors of sexual intercourse and readiness for sexual intercourse were examined across a range of sociodemographic measures. Overall, 17.7 % (95 % CI 16.7 %, 18.9 %) of participants reported having had sexual intercourse in the last year, with more girls than boys reporting sexual experience (risk ratio 1.30, 95 % CI 1.15, 1.47). Of these, one-third of both male and female were classed as unready because they were unwilling, lacking in autonomy, felt regret or had not used contraception. There was strong evidence of social patterning for sexual activity with higher rates for young people from poorer homes, with lower social class, and with younger, less educated mothers. In contrast, among 860 young people who had had sexual intercourse, there was no clear evidence of associations between social factors and sexual readiness. The lack of social patterning in sexual readiness supports the provision of comprehensive education to develop life skills for adolescents across all social groups.

Health-related quality of life in rural children living in four European countries: the GABRIEL study

OBJECTIVE Measuring children's health-related quality of life (HRQOL) is of growing importance given increasing chronic diseases. By integrating HRQOL questions into the European GABRIEL study, we assessed differences in HRQOL between rural farm and non-farm children from Germany, Austria, Switzerland and Poland to relate it to common childhood health problems and to compare it to a representative, mostly urban German population sample (KIGGS). METHODS The parents of 10,400 school-aged children answered comprehensive questionnaires including health-related questions and the KINDL-R questions assessing HRQOL. RESULTS Austrian children reported highest KINDL-R scores (mean: 80.9; 95 % CI [80.4, 81.4]) and Polish children the lowest (74.5; [73.9, 75.0]). Farm children reported higher KINDL-R scores than non-farm children (p = 0.002). Significantly lower scores were observed in children with allergic diseases (p < 0.001), with sleeping difficulties (p < 0.001) and in overweight children (p = 0.04). The German GABRIEL sample reported higher mean scores (age 7-10 years: 80.1, [79.9, 80.4]; age 11-13 years: 77.1, [74.9, 79.2]) compared to the urban KIGGS study (age 7-10 years: 79.0, [78.7-79.3]; age 11-13 years: 75.1 [74.6-75.6]). Socio-demographic or health-related factors could not explain differences in HRQOL between countries. CONCLUSIONS Future increases in chronic diseases may negatively impact children's HRQOL.