Simultaneous inference in structured additive conditional copula regression models: a unifying Bayesian approach

While most regression models focus on explaining distributional aspects of one single response variable alone, interest in modern statistical applications has recently shifted towards simultaneously studying multiple response variables as well as their dependence structure. A particularly useful tool for pursuing such an analysis are copula-based regression models since they enable the separation of the marginal response distributions and the dependence structure summarised in a specific copula model. However, so far copula-based regression models have mostly been relying on two-step approaches where the marginal distributions are determined first whereas the copula structure is studied in a second step after plugging in the estimated marginal distributions. Moreover, the parameters of the copula are mostly treated as a constant not related to covariates and most regression specifications for the marginals are restricted to purely linear predictors. We therefore propose simultaneous Bayesian inference for both the marginal distributions and the copula using computationally efficient Markov chain Monte Carlo simulation techniques. In addition, we replace the commonly used linear predictor by a generic structured additive predictor comprising for example nonlinear effects of continuous covariates, spatial effects or random effects and furthermore allow to make the copula parameters covariate-dependent. To facilitate Bayesian inference, we construct proposal densities for a Metropolis–Hastings algorithm relying on quadratic approximations to the full conditionals of regression coefficients avoiding manual tuning. The performance of the resulting Bayesian estimates is evaluated in simulations comparing our approach with penalised likelihood inference, studying the choice of a specific copula model based on the deviance information criterion, and comparing a simultaneous approach with a two-step procedure. Furthermore, the flexibility of Bayesian conditional copula regression models is illustrated in two applications on childhood undernutrition and macroecology.     

“Privacy please!” The public to private decision in Germany

This paper examines the characteristics and motives of German public to private transactions (PTPs) between 1996 and 2004. We empirically test hypotheses regarding free cash flow, leverage potential, ownership concentration, a stock market quotation’s decreasing benefits, limited capital market efficiency, and dividends. The results illustrate that, in the relevant period, the going private companies originated from mature industries and that, in the three financial years preceding the PTP transaction, their sales figures showed slow growth. Furthermore, the capital structure had potential for further leveraging and the companies were marked by low profitability. Trading in the PTPs’ shares was very sluggish, which limited the efficiency of the capital markets. Ultimately, we do not find evidence supporting the free cash flow hypothesis regarding German firms.     

Primary Care Screening of Depression and Treatment Engagement in a University Health Center: A Retrospective Analysis

Abstract

Objectives: This retrospective study analyzed a primary care depression screening initiative in a large urban university health center. Depression detection, treatment status, and engagement data are presented. Participants: Participants were 3,713 graduate and undergraduate students who presented consecutively for primary care services between January and April 2006. Methods: A standardized 2-tiered screening approach for an inception cohort of students utilizing primary services. Primary care providers were trained to triage students with depressive symptoms. Results: Six percent of participants had clinically significant depressive symptoms (CSD). Severe depressive symptoms were found in less than 1.0% of participants. Male rates of severe depressive symptoms were more than double that of females. Only 35.7% of untreated depressed participants started treatment within 30 days following identification. Conclusions: Systematic primary care depression screening in a college health center is a promising approach to identify untreated students with depression. More study is needed to improve rates of treatment engagement.     

Impediments to Academic Performance of Bisexual College Students

Abstract

Objective: To investigate health-related impediments to academic success for bisexual college students. Participants: Respondents to the Fall 2011 American College Health Association–National College Health Assessment II (ACHA-NCHA II) survey who self-identified as bisexual, heterosexual, gay, or lesbian. Methods: Secondary analyses of the ACHA-NCHA II data were conducted to compare the experiences of bisexual students with both each other (bisexual men and bisexual women) and those of their nonbisexual counterparts (heterosexual and gay/lesbian students). Results: Bisexual college students are a large and distinct sexual minority group. On all measures with the exception of discrimination, bisexual college students reported the strongest threats to academic success of all sexual orientations. Conclusions: Given the unique impediments for bisexual students, college health professionals should consider specifically addressing needs for bisexual students, bisexual women in particular.     

Predictors of sexual risk‐taking among new drug users

We studied predictors of HIV‐related sexual risk‐taking among individuals who initiated the use of heroin or metham‐phetamine during the past 5 years. Many studies have linked drug use to HIV risk, but most research has been based on samples of users with long‐established drug careers. We conducted face‐to‐face interviews with 153 adult new drug users in Atlanta, GA. Drug use was not a statistically significant predictor of sexual risk, but gender, age, race, homelessness status, childhood neglect, level of paranoia, and level of (dys)functionality in handling disagreements were associated with frequency of sexual risk‐taking. We discuss the need to incorporate new drug users in HIV/AIDS and other health‐related prevention and intervention programs.     

Self-generated motives for gambling in two population-based samples of gamblers

In the present study, self-generated responses to a question regarding reasons for gambling from two epidemiological surveys were combined and placed into another earlier motivational model for alcohol use, adapted for gambling. Of the 3601 reasons, 954 could be categorised into the model's categories: (a) coping motives (internal, negative reinforcement); (b) enhancement motives (internal, positive reinforcement); and (c) social motives (external, positive reinforcement). Results indicate that coping gamblers experienced greater gambling severity and psychopathology, enhancement gamblers were most likely to gamble while intoxicated and social gamblers were more likely to choose socially-related gambling. An examination of remaining motives suggests additional categories may be warranted – specifically financial and charitable reasons. These findings offer some support for the model; however, it may need to be expanded to account for other motives. The study highlights the advantages and limitations of using self-generated reasons to study gambling motivation.     

Toward Multifunctional Landscapes: Cross‐Sectional Analysis of Management Priorities in New York's Northern Forest*

Abstract Globalization and shifting societal relations with nature generate change and conflicting politics in rural areas of industrialized nations. In this context, multifunctionality has emerged as an important policy logic to stabilize commodity production while encouraging amenity‐based development and the production of ecological services. We argue that the concept of multifunctionality, while originally derived from agricultural trade negotiations, is applicable to integrated rural development challenges in forested landscapes. To understand and realize a progressive form of multifunctionality, information is needed on local actors' land use priorities and the factors that shape their interests. Based on data from a 2004 mail survey (183 valid responses; 61% response rate) in two northern NY communities, one amenity‐oriented and the other timber‐dependent, we conducted an empirical analysis of actors' forest management priorities. Factor analysis allowed us to identify four distinct orientations to management: ecological orientation, recreation orientation, large‐scale timber orientation, and small‐scale industry orientation. GLM regression allows us to assess how economic context, professional roles, and personal lifestyle considerations affect individuals' adherence to these distinct orientations. The commonalities and differences we observe among sets of actors engaged in forest policy and management point to constraints to realization of multifunctionality and to opportunities to develop richer conceptual tools to probe contemporary transitions.     

Women’s Reproductive Choices in Australia: Mapping Federal and State/Territory Policy Instruments Governing Choice

The reproductive choices available to women, and the consequences of those choices, exist within the broader policy context whereby policy is influenced by pervasive ideologies of women’s roles in society and the family. Women’s reproductive rights and their resulting consequences are governed by policy at the federal and state/territory levels within Australia yet little is known about the number or scope of these policies. This study aimed to systematically search and map Australian policy to identify the number and scope of policies governing women’s reproductive choices and their consequences, including how policy interprets the role of women in society through their reproductive choices. The systematic search identified 147 Australian policies in 2013. The mapping of the policies identified common themes that drive policy agenda impacting women’s reproductive choices, including those where the focus is promoting motherhood and/or children, providing economic incentives, regulating reproduction, or a broader health focus. These policy agendas simultaneously construct and are shaped by the context in which women’s reproductive choices and impacts occur. Women’s reproductive choices are highly politicised and regulated, impacting women’s position within society.