A note On outlier sensitivity of Sliced Inverse Regression

Sliced Inverse Regression (SIR) is a promising technique for the purpose of dimension reduction. Several properties of this method have been examined already, but little attention has been paid to robustness aspects. In this article, we focus on the sensitivity of SIR to outliers and show in what sense and how severely SIR can be influenced by outliers in the data.     

The flood algorithm—a multivariate, self-organizing-map-based, robust location and covariance estimator

Self-organizing maps (SOMs) introduced by Kohonen (Biol. Cybern. 43(1):59–69, 1982) are well-known in the field of artificial neural networks. The way SOMs are performing is very intuitive, leading to great popularity and numerous applications (related to statistics: classification, clustering). The result of the unsupervised learning process performed by SOMs is a non-linear, low-dimensional projection of the high-dimensional input data, that preserves certain features of the underlying data, e.g. the topology and probability distribution (Lee and Verleysen in Nonlinear Dimensionality Reduction, Springer, 2007; Kohonen in Self-organizing Maps, 3rd edn., Springer, 2001).     

Sind türkischstämmige Kinder beim Schulstart im Nachteil?

This article analyses the effect of a Turkish migration background on the risk of delayed school entry. Besides the social background, we control general and host country specific individual skill levels of the child. We assume that children’s individual skills can affect the school entry decision both directly and indirectly. We use data of the longitudinal project “Preschool education and educational careers among migrant children” (German acronym ESKOM VG). About 500 German parents and 500 Turkish-origin parents were interviewed in this study. Standardized developmental tests were conducted with the children in various domains. We find that the Turkish migration background does not have an independent effect on the probability of delayed school entry when the social background and individual skills are controlled. We furthermore disentangle the indirect effect of the Turkish migration background using the khb method. The results indicate that mathematic skills and cultural knowledge have the largest contribution on this indirect effect.     

Talks Between Teachers

Qualitative Sociology -     

Graduate education

Graduate training in sociology is an uneasy compromise between teaching new sociologists what would be good for them to know and doing what a graduate department’s various constituencies demand. Instead of worrying about teaching a core of materials, the graduate faculty instead should teach students what they know, and try to help students learn what they want to know, as they come to define that during the course of their studies. Instead of requiring courses, faculty should develop a continuing dialogue with students and incorporate them, formally and informally, in their work. His most recent book isDoing Things Together (1986).     

Measures of return under uncertainty

The purpose of this paper is twofold. First a methodlogy for contrasting accounting and economic measures of return will be briefly described. This methodology overcomes many of the limitations of previous comparisons of these measures. Second, the results of applying the new methodology to an area of current accounting concern—adjusting for price level changes—will be presented.     

Suitability of physical activity questionnaires for older adults in fall-prevention trials: a systematic review

The purpose of the study was to identify physical activity questionnaires for older adults that might be suitable outcome measures in clinical trials of fall-injury-prevention intervention and to undertake a systematic quality assessment of their measurement properties. PubMed, CINAHL, and PsycINFO were systematically searched to identify measurements and articles reporting the methodological quality of relevant measures. Quality extraction relating to content, population, reliability, validity, responsiveness, acceptability, practicality, and feasibility was undertaken. Twelve outcome measures met the inclusion criteria. There is limited evidence about the measures' properties. None of the measures is entirely satisfactory for use in a large-scale trial at present. There is a need to develop suitable measures. The Stanford 7-day Physical Activity Recall Questionnaire and the Community Health Activities Model Program for Seniors questionnaire might be appropriate for further development. The results have implications for the designs of large-scale trials investigating many different geriatric syndromes.     

ESTIMATING POPULATION SIZE FROM CAPTURE-RECAPTURE EXPERIMENTS IN CONTINUOUS TIME

Estimates for the size of a closed population are given for multiple recapture studies in continuous time. The estimates are derived by a method of moments for martingales. An estimate and associated standard error of the population size are derived for a homogeneous population when the capture rates are permitted to depend on time in an unspecified manner. Corresponding results are obtained when the capture rates vary among individuals as well. Explicit expressions are given for these estimates and standard errors which involve only simple computation.