Two-stage Huber estimation

In this paper we propose a new robust estimator in the context of two-stage estimation methods directed towards the correction of endogeneity problems in linear models. Our estimator is a combination of Huber estimators for each of the two stages, with scale corrections implemented using preliminary median absolute deviation estimators. In this way we obtain a two-stage estimation procedure that is an interesting compromise between concerns of simplicity of calculation, robustness and efficiency. This method compares well with other possible estimators such as two-stage least-squares (2SLS) and two-stage least-absolute-deviations (2SLAD), asymptotically and in finite samples. It is notably interesting to deal with contamination affecting more heavily the distribution tails than a few outliers and not losing as much efficiency as other popular estimators in that case, e.g. under normality. An additional originality resides in the fact that we deal with random regressors and asymmetric errors, which is not often the case in the literature on robust estimators.     

Understanding Canonical Correlation through the General Linear Model and Principal Components

Canonical correlation has been little used and little understood, even by otherwise sophisticated analysts. An alternative approach to canonical correlation, based on a general linear multivariate model, is presented. Properties of principal component analysis are used to help explain the method. Standard computational methods for full rank canonical correlation, techniques for canonical correlation on component scores, and canonical correlation with less than full rank are discussed. They are seen to be essentially equivalent when the model equation for canonical correlation on component scores is presented. The two approaches to less than full rank situations are equivalent in some senses, but quite different in usefulness, depending on the application. An example dataset is analyzed in detail to help demonstrate the conclusions.     

Quantitative Risk Assessment of Listeria monocytogenes in French Cold-Smoked Salmon: II. Risk Characterization

A model for the assessment of exposure to Listeria monocytogenes from cold-smoked salmon consumption in France was presented in the first of this pair of articles (Pouillot et al ., 2007, Risk Analysis, 27:683–700). In the present study, the exposure model output was combined with an internationally accepted hazard characterization model, adapted to the French situation, to assess the risk of invasive listeriosis from cold-smoked salmon consumption in France in a second-order Monte Carlo simulation framework. The annual number of cases of invasive listeriosis due to cold-smoked salmon consumption in France is estimated to be 307, with a very large credible interval ([10; 12,453]), reflecting data uncertainty. This uncertainty is mainly associated with the dose-response model. Despite the significant uncertainty associated with the predictions, this model provides a scientific base for risk managers and food business operators to manage the risk linked to cold-smoked salmon contaminated with L. monocytogenes. Under the modeling assumptions, risk would be efficiently reduced through a decrease in the prevalence of L. monocytogenes or better control of the last steps of the cold chain (shorter and/or colder storage during the consumer step), whereas reduction of the initial contamination levels of the contaminated products and improvement in the first steps of the cold chain do not seem to be promising strategies. An attempt to apply the recent risk-based concept of FSO (food safety objective) on this example underlines the ambiguity in practical implementation of the risk management metrics and the need for further elaboration on these concepts.     

The proximity filter: the effect of distance on media coverage of the Christchurch mosque attacks

ABSTRACT

The intense media coverage in New Zealand and Australia of the Christchurch mosque attacks exhibited significant disparity in editorial decision-making between the two countries. This research interrogates the different approaches taken in newsrooms and how these differences were manifested in broadcasts and publications. New Zealand media were focused largely on empathetic coverage of victims and resisted the alleged gunman's attempts to publicise his cause while their Australian counterparts showed no such reluctance and ran extended coverage of the alleged perpetrator, along with material ruled objectionable in New Zealand. It finds the editorial focus in each case exhibits the effect of proximity, identified in literature on empirical ethical decision-making as a factor in applied ethicality. The authors conclude that a proximity filter was used by New Zealand media who identified the victims as part of their own community, but the events of 15 March 2019 were seen as ‘foreign’ by Australian journalists who used perceived distance as justification for extremely graphic content.     

Adolescent Religiosity and School Contexts*

Objectives. Schools serve as the primary social organizations for adolescents, structuring their lives and conveying a variety of skills, norms, and values, but relatively little is known about how schools influence the development of religious belief, attitudes, and behavior during adolescence. We explore how schools' religious norms, coupled with adolescents' pursuit of social status through conformity, affect public and private religiosity. Methods. Employing data from the National Longitudinal Study of Adolescent Health (Add Health), we use multilevel modeling to examine the impact of school context on adolescents' public and private religiosity. Results. We find that school norms are influential in shaping both public religious expressions and private devotional activities, but pursuit of social status is a motivation for change across religious contexts only of public religious activity. The effect of social status as a motivator of religious change was moderated by the strength of the adolescent's identification with the school, especially for private religiosity. Conclusions. Schools play a key role in the social development of adolescents, and students' religious beliefs and behaviors are influenced systematically and observably by the type of religious climate within the school.     

Regularization through variable selection and conditional MLE with application to classification in high dimensions

It is often the case that high-dimensional data consist of only a few informative components. Standard statistical modeling and estimation in such a situation is prone to inaccuracies due to overfitting, unless regularization methods are practiced. In the context of classification, we propose a class of regularization methods through shrinkage estimators. The shrinkage is based on variable selection coupled with conditional maximum likelihood. Using Stein's unbiased estimator of the risk, we derive an estimator for the optimal shrinkage method within a certain class. A comparison of the optimal shrinkage methods in a classification context, with the optimal shrinkage method when estimating a mean vector under a squared loss, is given. The latter problem is extensively studied, but it seems that the results of those studies are not completely relevant for classification. We demonstrate and examine our method on simulated data and compare it to feature annealed independence rule and Fisher's rule.     

The Good, the Bad, and the Regulator: An Experimental Test of Two Conditional Audit Schemes

Conditional audit rules are designed to achieve regulatory compliance with fewer inspections than required by random auditing. A regulator places individuals into audit pools that differ in probability of audit or severity of fine and specifies transition rules between pools. Future pool assignment is conditional on current audit results. We conduct an experiment to compare two specific schemes—Harrington's Past-Compliance Targeting and Friesen's Optimal Targeting—against random auditing. We find a production possibility frontier between compliance and minimizing inspections. Optimal targeting generates the lowest inspection rates as predicted, but random auditing the highest compliance. Past-compliance targeting is intermediate.     

Of Sex and Romance: Late Adolescent Relationships and Young Adult Union Formation

To better understand the social factors that influence the diverse pathways to family formation young adults experience today, this research investigates the association between opposite‐gender relationships during late adolescence and union formation in early adulthood. Using data from the first and third waves of the Add Health (n= 4,911), we show that, for both men and women, there is continuity between adolescent and adult relationship experiences. Those involved in adolescent romantic relationships at the end of high school are more likely to marry and to cohabit in early adulthood. Moreover, involvement in a nonromantic sexual relationship is positively associated with cohabitation, but not marriage. We conclude that the precursors to union formation patterns in adulthood are observable in adolescence.