Cross-validating fit and predictive accuracy of nonlinear quantile regressions

The paper proposes a cross-validation method to address the question of specification search in a multiple nonlinear quantile regression framework. Linear parametric, spline-based partially linear and kernel-based fully nonparametric specifications are contrasted as competitors using cross-validated weighted L ₁-norm based goodness-of-fit and prediction error criteria. The aim is to provide a fair comparison with respect to estimation accuracy and/or predictive ability for different semi- and nonparametric specification paradigms. This is challenging as the model dimension cannot be estimated for all competitors and the meta-parameters such as kernel bandwidths, spline knot numbers and polynomial degrees are difficult to compare. General issues of specification comparability and automated data-driven meta-parameter selection are discussed. The proposed method further allows us to assess the balance between fit and model complexity. An extensive Monte Carlo study and an application to a well-known data set provide empirical illustration of the method.     

海因里希·伯尔与文学现代性

本文通过对伯尔的代表作《女士及众生相》的分析,证明伯尔并非一个传统意义上的现实主义作家,《女士及众生相》通过对现实主义形式的借用和过度实施完成的恰恰是对传统现实主义叙事模式的侵蚀,伯尔的小说完全可以接入艺术自主的美学现代性传统。     

Loss prediction in a linear model under a linear constraint

Loss reserving is an important subject of actuarial mathematics. It aims at the prediction of future losses caused by claims which have incurred in the past but have not yet been closed. The problem of predicting such losses is particularly important in liability insurance. In the present paper we study conjoint prediction of paid and incurred losses in a linear model with a linear constraint which is intended to reduce the gap between the predictors of ultimate paid and incurred losses. We thus present an application to actuarial mathematics of the general result established by Kloberdanz and Schmidt (AStA Adv. Stat. Anal. 92:207–215, 2008).     

The impact of different fit dimensions on spillover effects in brand alliances

Brand alliances represent a popular business strategy in many industries, because firms hope to evoke positive consumer evaluations of both the alliance’s product and the partner brands. However, extant research offers mixed findings regarding the effects of a brand alliance on its partner brands (i.e., spillover effects). In response, this study separates spillover effects into the effects of the alliance product on the partner brands (brand alliance effects) and the effects between partner brands (brand contrast effects), while also noting the potential moderating impact of perceived attitude- and product-based fit between partner brands on resulting spillover effects. Two experimental studies consistently reveal the existence of brand contrast effects; furthermore, the partner brand’s attitude-based fit reduces undesired brand contrast effects and positively moderates spillover effects in brand alliances, whereas product-based fit does not. Therefore, a third study identifies relevant drivers of partner brand’s attitude-based fit for different brand alliances (i.e., co-branding, ingredient branding, and joint advertising). The findings have notable implications for the design and management of brand alliances.     

Untersuchungen zur Ursache unterschiedlicher Ergebnisse sehr ähnlicher Viktimisierungssurveys

Two very similar surveys on victimization were conducted in 1997. Although the target population, design, operationalization, field organization and part of the interviewer staff were nearly identical, the resulting estimates of victimization status were remarkably different. Neither correction for design effects nor the use of appropriate weighting eliminated this statistically significant difference. Several methodological details of the surveys are considered as explanations. Quantitative estimates of their possible effects are given. The interviewer working conditions are identified as the most likely source of the discrepancy. This result emphasizes the importance of these seldom documented and even more rarely controlled fieldwork details.     

Who says what to whom: Why messengers and citizen beliefs matter in social policy framing

Issue frames (i.e., the thematic slants that elites use to structure issue debates) have been shown to alter how citizens think about social policy issues. However, support for a social policy issue not only depends on how the issue is framed, but also on the source or “messenger” associated with the frame. For the most part, issue frames have been faceless and research has failed to consider how characteristics of the frame's messenger such as expertise and trustworthiness influence citizens. The present study examines the influence that gun control frames and their messengers have on a variety of dependent variables. The results show that source cues moderate the impact of the frames, even when controlling for other variables. Specifically, credible sources significantly augment the effects of a frame while spokespersons with a perceived bias tend to weaken the frame's arguments and impact. These results point toward a more complex framing theory: public support for social issues depends on how the issue is framed, as well as who presents the message.     

Beyond growth: new alliances for socio-ecological transformation in Austria

ABSTRACT

Trade unions and environmental movements are often seen as political opponents most prominently discussed in the form of the ‘jobs vs. environment dilemma’. Based on historical examples of the conflict relations between trade unions and environmental groups in the Austrian energy sector, this paper showcases how the relationship between the two groups has changed from enmity to first attempts at alliance building. Drawing from analysis of union documents and problem-centred interviews conducted with Austrian unionists, it shows that newly emerging alliances between unions and environmental movements contain the seeds for a broad societal movement that can help overcome the paradigm of growth and actively engage in the creation of policies that support a social–ecological transformation.     

Explaining the heterogeneity of the leadership-innovation relationship: Ambidextrous leadership

The authors review and meta-analytically integrate the existing literature on leadership and innovation to show a complex and inconsistent picture of this relationship. Current research has mostly neglected the complex nature of innovation processes that leads to changing requirements within these processes. The main requirements of innovation are exploration and exploitation as well as a flexibility to switch between those two activities. The authors propose an ambidexterity theory of leadership for innovation that specifies two complementary sets of leadership behavior that foster exploration and exploitation in individuals and teams — opening and closing leader behaviors, respectively. We call this ambidextrous leadership because it utilizes opening and closing leader behaviors and switches between them to deal with the ever-changing requirements of the innovation process. Routes to ambidextrous leadership and opportunities for future research on leadership and innovation are discussed.