On the $$L_p$$ norms of kernel regression estimators for incomplete data with applications to classification

We consider kernel methods to construct nonparametric estimators of a regression function based on incomplete data. To tackle the presence of incomplete covariates, we employ Horvitz–Thompson-type inverse weighting techniques, where the weights are the selection probabilities. The unknown selection probabilities are themselves estimated using (1) kernel regression, when the functional form of these probabilities are completely unknown, and (2) the least-squares method, when the selection probabilities belong to a known class of candidate functions. To assess the overall performance of the proposed estimators, we establish exponential upper bounds on the \(L_p\) norms, \(1\le p<\infty \), of our estimators; these bounds immediately yield various strong convergence results. We also apply our results to deal with the important problem of statistical classification with partially observed covariates.     

Ethnicity-of-Interviewer Effects Among Mexican-Americans and Anglos

A general interpersonal deference explanation for ethnicity-of-interviewereffects is tested with a random survey of Anglos and Mexican-Americansusing Anglo and Hispanic interviewers and questions relatedto Mexican-American culture. The theory was most clearly supportedamong Anglo respondents. The mixed support found among Mexican-Americanssuggests that deference is strongest when questions deal withthe culture represented by the interviewer, not the respondent.     

The Social Work Assessment Tool (SWAT)

This paper reports on the last of three National Hospice and Palliative Care Organization initiatives to move hospice and palliative care social workers into the patient/family outcomes arena: the development of the Social Work Assessment Tool. The experience of a team of practitioners and researchers is described, including results of two pilot studies and subsequent SWAT revisions. The major focus is on the current model performance improvement project, in which 19 social workers from 14 hospice and palliative care programs used the SWAT with 101 patients and 81 primary caregivers for a median of 44 days. Quantitative analysis indicated significant improvement in SWAT scores for patients from the first to the second social work visit (t = -2.60, df = 47, p .01). Qualitative interviewing of the social workers indicated some lack of readiness in the field to conduct quantitative outcomes measurement. Additional measures are needed in addition to the SWAT, including qualitative measures, and measures of mezzo and macro practice. Participants indicated that the SWAT was appropriate for use with economically and culturally diverse clients.     

Racial differences in infant mortality by cause of death: The impact of birth weight and maternal age

This research examines racial disparities in infant mortality, overall and separately according to cause of death. Using linked birth and death records for the 1975 cohort of live births in Florida, racial differences are initially described and then explained statistically as a product of the distribution of births by birth weight and maternal age. The impact of birth weight is more pronounced than is the effect of maternal age. The analysis suggests the potential utility of examining infant mortality separately by cause of death. Based on the findings, we argue for systematic research focused on factors affecting birth weight.     

Understanding the Self Through Siblings: Self‐awareness Mediates the Sibling Effect on Social Understanding

This study addresses the role of sibling influence on social understanding in the second year of life, in a sample of families living in New Zealand who identify as Pacific Island (N = 43). We tested toddlers at 20 and 26 months on social understanding tasks, as well as their levels of self‐awareness on the Stipek self‐concept questionnaire. We hypothesized that the presence of siblings provides a rich resource from which to learn about the mind. There were significant differences in children's social understanding and self‐awareness as a function of having older siblings. Further analyses revealed that the relation between older siblings and social understanding was mediated by toddlers' level of self‐awareness.     

Journalism and Globalization

Like every other social practice, journalism cannot now fully be understood apart from globalization. As part of a larger platform of communication media, journalism contributes to this experience of the world-as-a-single-place and thus represents a key component in these social transformations, both as cause and outcome. These issues at the intersection of journalism and globalization define an important and growing field of research, particularly concerning the public sphere and spaces for political discourse. In this essay, I review this intersection of journalism and globalization by considering the communication field’s approach to ‘media globalization’ within a broader interdisciplinary perspective that mixes the sociology of globalization with aspects of geography and social anthropology. By placing the emphasis on social practices, elites, and specific geographical spaces, I introduce a less media-centric approach to media globalization and how journalism fits into the process. Beyond ‘global village journalism,’ this perspective captures the changes globalization has brought to journalism.     

Estimation of factor-augmented panel regressions with weakly influential factors

The use of factor-augmented panel regressions has become very popular in recent years. Existing methods for such regressions require that the common factors are strong, an assumption that is likely to be mistaken in practice. Motivated by this, the current article offers an analysis of the effect of weak, semi-weak, and semi-strong factors on two of the most popular estimators for factor-augmented regressions, namely, principal components (PC) and common correlated effects (CCE).     

Literatur / Books

Buchbesprechungen / Book reviews

Literatur / Books