A comparison of heritage language ideologies in interaction

In Canada, the notion of a heritage language ideology is often conceived of as a natural by‐product of official multiculturalism. By contrast, Germany has long struggled with its status as a multilingual and multicultural country. By comparing two corpora of interviews with immigrants to each of these two countries (Canadians of German heritage and Germans of Vietnamese heritage), this paper aims to explore to what extent these different language ideologies are reconstructed in the interviews. It will be argued that the interviewees construct different sociolinguistic spaces and take up different positions within them in terms of centre and periphery. Our analysis shows that the German‐Canadian interviewees construct public sociolinguistic spaces in which they position themselves as German even when they do not have an active knowledge of their heritage language. By contrast, despite the monolingual habitus in Germany, the German‐Vietnamese respondents endorse a heritage language ideology; the space they claim for speaking Vietnamese, however, is restricted to private or family conversations.     

Approximation of distributions by using the Anderson Darling statistic

ABSTRACT

In practice, it is often not possible to find an appropriate family of distributions which can be used for fitting the sample distribution with high precision. In these cases, it seems to be opportune to search for the best approximation by a family of distributions instead of an exact fit. In this paper, we consider the Anderson–Darling statistic with plugged-in minimum distance estimator for the parameter vector. We prove asymptotic normality of the Anderson–Darling statistic which is used for a test of goodness of approximation. Moreover, we introduce a measure of discrepancy between the sample distribution and the model class.     

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).     

Language attitudes in interaction1

This paper discusses the observation of language attitudes in interaction and argues that these approaches provide invaluable insights for the study of language attitudes. In the first half of the paper, the three different kinds of discourse‐based methods of analysis that scholars have used to analyse language attitudes (content‐based approaches, turn‐internal semantic and pragmatic approaches, and interactional approaches) are discussed. In the second half, then, the third of these approaches is used to illustrate such an analysis with four stretches of conversation in different contexts. In the end, the argument is put forward that discourse‐based approaches in general and interactional approaches in particular should be viewed as at least as fundamental to language attitude research as more commonly used quantitative methods of analysis, since the former can provide the researcher with insights that the latter do not.     

A generalization of the Kaplan-Meier estimator to Harris-recurrent Markov chains

In this paper we study an extension of the Kaplan Meier estimator to the case of dependent failure times. We assume that the failure times follow a Harris recurrent Markov Chain. We prove strong convergence of the estimator and weak convergence to a Gaussian process.     

The returns to continuous training in Germany: new evidence from propensity score matching estimators

The present paper examines the wage effects of continuous training programs using individual-level data from the German Socio Economic Panel (GSOEP). In order to account for selectivity in training participation we estimate average treatment effects (ATE and ATT) of general and firm-specific continuous training programs using several state-of-the-art propensity score matching (PSM) estimators. Additionally, we also apply a combined matching difference-in-differences (MDiD) estimator to account for unobserved individual characteristics (e.g. motivation, ability). While the estimated ATE and ATT for general training are significant ranging between about 4 and 7.5%, the corresponding wage effects of firm-specific training are mostly insignificant. Using the more appropriate MDiD estimator, however, we find a more precise and highly significant wage effect of about 5–6%, though only for general training and not for firm-specific training. These results are consistent with standard human capital theory insofar as general training is associated with larger wage increases than firm-specific training. Furthermore, we conclude that firms may intend to use specific training to adjust to new job requirements, while career-relevant changes may be conditioned to general training.