Diagnostics for multivariate imputations

Summary.  We consider three sorts of diagnostics for random imputations: displays of the completed data, which are intended to reveal unusual patterns that might suggest problems with the imputations, comparisons of the distributions of observed and imputed data values and checks of the fit of observed data to the model that is used to create the imputations. We formulate these methods in terms of sequential regression multivariate imputation, which is an iterative procedure in which the missing values of each variable are randomly imputed conditionally on all the other variables in the completed data matrix. We also consider a recalibration procedure for sequential regression imputations. We apply these methods to the 2002 environmental sustainability index, which is a linear aggregation of 64 environmental variables on 142 countries.     

To be or not to be: the effects of aging stereotypes on the will to live

This study examined whether stereotypes of aging might contribute to decisions the elderly make about when to die. Old and young participants (N=64) were subliminally primed with either negative or positive stereotypes of old age and then responded to hypothetical medical situations involving potentially fatal illnesses. Consistent with our prediction, the aged participants primed with negative stereotypes tended to refuse life-prolonging interventions, whereas the old participants primed with positive age stereotypes tended to accept the interventions. This priming effect did not emerge among the young participants for whom the stereotypes were less relevant. The results suggest that societally-transmitted negative stereotypes of aging can weaken elderly people's will to live.     

On the definition and varieties of attitude and wellbeing

In a paper published in Social Indicators Research in 1975, Levy and Guttman proved the proposition that wellbeing is a special case of attitude. They exploited this result for establishing a certain lawfulness for wellbeing behavior, and built on this for revealing more specialized features of the lawfulness. Five years later in this same journal, Andrews and McKennell published two papers on exactly the same proposition, but without proving it, without building on known relevant lawfulness of attitudinal behavior, and without mentioning (neither positively nor negatively) the Levy-Guttman work. The Andrews-McKennell papers are shown to be a retrogression rather than a scientific advance, being conceptually confused, using inappropriate data analysis, and dealing with arbitrary assumptions which are either untested, untestable or false.     

Integrating risk assessment and life cycle assessment: a case study of insulation.

Increasing residential insulation can decrease energy consumption and provide public health benefits, given changes in emissions from fuel combustion, but also has cost implications and ancillary risks and benefits. Risk assessment or life cycle assessment can be used to calculate the net impacts and determine whether more stringent energy codes or other conservation policies would be warranted, but few analyses have combined the critical elements of both methodologies In this article, we present the first portion of a combined analysis, with the goal of estimating the net public health impacts of increasing residential insulation for new housing from current practice to the latest International Energy Conservation Code (IECC 2000). We model state-by-state residential energy savings and evaluate particulate matter less than 2.5 microm in diameter (PM2.5), NOx, and SO2 emission reductions. We use past dispersion modeling results to estimate reductions in exposure, and we apply concentration-response functions for premature mortality and selected morbidity outcomes using current epidemiological knowledge of effects of PM2.5 (primary and secondary). We find that an insulation policy shift would save 3 x 10(14) British thermal units or BTU (3 x 10(17) J) over a 10-year period, resulting in reduced emissions of 1,000 tons of PM2.5, 30,000 tons of NOx, and 40,000 tons of SO2. These emission reductions yield an estimated 60 fewer fatalities during this period, with the geographic distribution of health benefits differing from the distribution of energy savings because of differences in energy sources, population patterns, and meteorology. We discuss the methodology to be used to integrate life cycle calculations, which can ultimately yield estimates that can be compared with costs to determine the influence of external costs on benefit-cost calculations.     

Beyond Residential Segregation: Mobility-Based Connectedness and Rates of Violence in Large Cities

A longstanding finding is that neighborhood racial segregation is linked to violence. In this paper, we look beyond neighborhoods of residence to consider the everyday mobility of urbanites in their daily rounds. Analyzing estimates of neighborhood mobility from largescale social media data in the 50 largest American cities, we find that residential segregation by race is not only associated with higher violence but also lower equitability of travel across neighborhoods and a lower concentration of visits to common hubs. Further, the interaction of equitable and concentrated mobility is significantly associated with rates of violence, controlling for both racial and income segregation, education, city size, and density. There is little evidence, however, that patterns of everyday mobility mediate the influence of residential racial segregation. Both dimensions of the structural connectedness of cities—one rooted in place of residence, and the other encompassing interneighborhood exposure based on travel throughout the metropolis—are implicated in violence.     

NASW’s Supervisory Leaders in Aging: An Acceptable and Feasible Model for Training and Supporting Social Work Supervisors

Social services are instrumental in addressing challenges associated with aging. Yet, practitioners report needing expanded gerontological knowledge and better supervision. The Supervisory Leaders in Aging (SLA) program of the National Association of Social Workers (NASW) was designed to improve gerontological services by strengthening supervision of the social service workforce. With support from the John A. Hartford Foundation, the program was adopted between 2015 and 2017 in four regions, where NASW trained 134 MSW supervisors who support 1200 social service staff, potentially enhancing the well-being of 264,000 clients annually. The SLA curriculum and in-person educational approach constitute a model for supervisor professional development. Quality improvement evaluations suggested feasibility of program adoption and acceptability. Participants rated each of 10 workshops and 97% agreed that instructors were effective, that knowledge was expanded, and that content was relevant, appropriate, and likely to be used. Implications of this model for enhancing supervisory practice are discussed.     

Practice-Specific Risk Perceptions and Self-Reported Food Safety Practices

The relationship between risk perception and risk avoidance is typically analyzed using self-reported measures. However, in domains such as driving or food handling, the validity of responses about usual behavior is threatened because people think about the situations in which they are self-aware, such as when they encounter a hazard. Indeed, researchers have often noted a divergence between what people say about their behavior and how they actually behave. Thus, in order to draw conclusions about risk perceptions and risk avoidance from survey data, it is important to identify particular cognitive elements, such as those measured by questions about risk and safety knowledge, risk perceptions, or information search behavior, which may be effective antecedents of self-reported safety behavior. It is also important to identify and correct for potential sources of bias that may exist in the data. The authors analyze the Food and Drug Administration's 1998 Food Safety Survey to determine whether there are consistent cognitive antecedents for three types of safe food practices: preparation, eating, and cooling of foods. An assessment of measurement biases shows that endogeneity of food choices affects reports of food preparation. In addition, response bias affects reports of cooling practices as evidenced by its relation to knowledge and information search, a pattern of cognitive effects unique to cooling practices. After correcting for these biases, results show that practice-specific risk perceptions are the primary cognitive antecedents of safe food behavior, which has implications for the design of effective education messages about food safety.     

The [R]Evolving Relationship Between Risk Assessment and Risk Management

In Science and Decisions: Advancing Risk Assessment, the National Research Council recommends improvements in the U.S. Environmental Protection Agency's approach to risk assessment. The recommendations aim to increase the utility of these assessments, embedding them within a new risk‐based decision‐making framework. The framework involves first identifying the problem and possible options for addressing it, conducting related analyses, then reviewing the results and making the risk management decision. Experience with longstanding requirements for regulatory impact analysis provides insights into the implementation of this framework. First, neither the Science and Decisions framework nor the framework for regulatory impact analysis should be viewed as a static or linear process, where each step is completed before moving on to the next. Risk management options are best evaluated through an iterative and integrative procedure. The extent to which a hazard has been previously studied will strongly influence analysts’ ability to identify options prior to conducting formal analyses, and these options will be altered and refined as the analysis progresses. Second, experience with regulatory impact analysis suggests that legal and political constraints may limit the range of options assessed, contrary to both existing guidance for regulatory impact analysis and the Science and Decisions recommendations. Analysts will need to work creatively to broaden the range of options considered. Finally, the usefulness of regulatory impact analysis has been significantly hampered by the inability to quantify many health impacts of concern, suggesting that the scientific improvements offered within Science and Decisions will fill an crucial research gap.