Bounding sample size projections for the area under a ROC curve

Studies of diagnostic tests are often designed with the goal of estimating the area under the receiver operating characteristic curve (AUC) because the AUC is a natural summary of a test's overall diagnostic ability. However, sample size projections dealing with AUCs are very sensitive to assumptions about the variance of the empirical AUC estimator, which depends on two correlation parameters. While these correlation parameters can be estimated from the available data, in practice it is hard to find reliable estimates before the study is conducted. Here we derive achievable bounds on the projected sample size that are free of these two correlation parameters. The lower bound is the smallest sample size that would yield the desired level of precision for some model, while the upper bound is the smallest sample size that would yield the desired level of precision for all models. These bounds are important reference points when designing a single or multi-arm study; they are the absolute minimum and maximum sample size that would ever be required. When the study design includes multiple readers or interpreters of the test, we derive bounds pertaining to the average reader AUC and the ‘pooled’ or overall AUC for the population of readers. These upper bounds for multireader studies are not too conservative when several readers are involved.     

Building the Sustainable Community: Is Social Capital the Answer?

Over the past decade, environmental sustainability has emerged as a prominent theme in the community development literature. In fact, the concept has become a standard feature of most economic and social development plans. Most models of sustainable community development stress the importance of widespread participation in the decision-making process. Unfortunately, community studies document numerous barriers to broad involvement and the high level of activeness envisioned by proponents of sustainable community development. In searching for ways to overcome these barriers, scholars and policymakers have embraced the idea that we can enhance efforts to create more sustainable communities by increasing the local stock of social capital. We examine this line of reasoning in light of what we view as the most important conceptual issues surrounding the relationship between social capital and sustainable community development. We conclude that before social capital is endorsed as a central component of public policy, much work remains to be done in terms of developing a more precise definition of the concept, situating it within extant theories of community, constructing better measures of social capital, documenting the activities and networks most important in building social capital, and gaining a better understanding of the forms of social capital that are most important in developing sustainable communities.     

Do Intentional Efforts to Improve Boards Really Work? The Views of Nonprofit CEOs

Commentators on the effectiveness of nonprofit boards of directors usually find them wanting in a number of ways and urge that they reform themselves. A study to ascertain how many boards actually accept this advice and intentionally attempt to change the way they operate also examines what changes boards attempt to make, what provokes their efforts, and what outcomes result.     

Identifying intraclass correlations necessitating hierarchical modeling

Hierarchical binary outcome data with three levels, such as disease remission for patients nested within physicians, nested within clinics are frequently encountered in practice. One important aspect in such data is the correlation that occurs at each level of the data. In parametric modeling, accounting for these correlations increases the complexity. These models may also yield results that lead to the same conclusions as simpler models. We developed a measure of intraclass correlation at each stage of a three-level nested structure and identified guidelines for determining when the dependencies in hierarchical models need to be taken into account. These guidelines are supported by simulations of hierarchical data sets, as well as the analysis of AIDS knowledge in Bangladesh from the 2011 Demographic Health Survey. We also provide a simple rule of thumb to assist researchers faced with the challenge of choosing an appropriately complex model when analyzing hierarchical binary data.     

Credit and authorship practices: educational and environmental influences

A survey on credit issues of academic chemists in U.S. Ph.D.-granting institutions was conducted. The respondents rated 15 criteria for authorship of scientific publications; core intellectual contributions received the highest ratings although making a single suggestion that was essential to the successful completion of the project was rated very low. Acquisition of data was also rated highly. The respondents rated eight potential influences on their own "policy" toward giving credit; doing what "seems to be the right thing" was the highest rated influence followed by graduate educational experiences; professional society or other responsible conduct of research (RCR) institutional policies were rated, by far, the lowest.     

High-Dimensional Sparse Factor Modeling: Applications in Gene Expression Genomics

We describe studies in molecular profiling and biological pathway analysis that use sparse latent factor and regression models for microarray gene expression data. We discuss breast cancer applications and key aspects of the modeling and computational methodology. Our case studies aim to investigate and characterize heterogeneity of structure related to specific oncogenic pathways, as well as links between aggregate patterns in gene expression profiles and clinical biomarkers. Based on the metaphor of statistically derived "factors" as representing biological "subpathway" structure, we explore the decomposition of fitted sparse factor models into pathway subcomponents and investigate how these components overlay multiple aspects of known biological activity. Our methodology is based on sparsity modeling of multivariate regression, ANOVA, and latent factor models, as well as a class of models that combines all components. Hierarchical sparsity priors address questions of dimension reduction and multiple comparisons, as well as scalability of the methodology. The models include practically relevant non-Gaussian/nonparametric components for latent structure, underlying often quite complex non-Gaussianity in multivariate expression patterns. Model search and fitting are addressed through stochastic simulation and evolutionary stochastic search methods that are exemplified in the oncogenic pathway studies. Supplementary supporting material provides more details of the applications, as well as examples of the use of freely available software tools for implementing the methodology.     

"EPISTEMOLOGY" AS A SEMANTIC POLLUTANT

Some family therapists, following Bateson, use "epistemology" in a peculiar, non-traditional way. Nothing is gained, and much appears to be lost, by this practice. Its main effect is to promote the idea that systemic and psychological modes of explanation are incompatible. In fact, a kind of cognitive psychology is implicit in what family therapists say about reframing. By appropriating the territory of psychology and calling it epistemology instead, family therapists merely pollute the semantic environment,¹ muddying the very things that theoretical terms are supposed to clear up.     

Distribution of neotropical migratory bird species across an urbanizing landscape

Urbanization leads to long-term modification of landscapes by habitat loss, fragmentation, and the creation of new habitats. Species’ distributions respond to these modifications of habitat availability, but the combination of parameters and scale at which habitat alteration most strongly influences species distributions is poorly understood. We evaluated responses of neotropical migratory birds, a group known to be sensitive to habitat modification, across a gradient of urbanization in the southeastern United States. Thirteen Breeding Bird Survey routes, each with 40 to 50 point counts, were used to quantify species richness across the gradient of urbanization extending from downtown areas of Columbus, GA to natural woodlands. Buffers of 100, 200, and 1000 m radii were constructed from remote images around each counting point to quantify land-use with the goal of evaluating land-use parameters and scales that best described spatial variation in migrant bird species richness. Within each buffer we quantified the proportion of each cover type and within the 1000 m buffers we included several configuration parameters. We used an information-theoretic approach to separate models whose predictor variables were land-use parameters. Because measures of landscape configuration were all correlated with urban cover, these variables were entered separately. In 2002, the best model was composed of large-scale urban cover (negative effect) and mid-scale mixed hardwoods (negative and positive effect) and transitional cover (negative and positive effect) as well as the interaction between the latter two terms (positive effect). In 2003, the best model was composed of weighted edge density (negative effect), mid-scale mixed hardwood cover (negative and positive effect) and large scale transitional cover (positive effect) and the interaction between mixed hardwoods and weighted edge density (positive effect). Our results indicate that large scale habitat attributes influence the local species richness of migrant birds more than smaller scales. These results also indicate that urbanization, through increased urban cover or increasing edge contrast, has strong negative effects on species richness. Our findings support the contention that the conservation value of small woodlots in urban settings may be minimal and suggest that conservation of migratory birds will be best achieved by giving higher priority to sites where urban cover is still low and by preserving large areas of “green space” in urbanizing landscapes. The negative influence of urban cover combined with relatively minor effects of non-urban habitats on distributions of neotropical migratory birds indicates that continued urbanization of landscapes is a serious concern for conservation efforts.     

Sodomites and police in Paris, 1715

The Archives of the Bastille in the Bibliothèque de l'Arsenal include hundreds of police reports, none of them currently available in English or even in print, about male same-sex relations in Paris during the first half of the eighteenth century. The documents translated here, interrogations of eight men arrested in 1715, provide information about networks and prostitution, age and class, surveillance and deception in the sodomitical subculture of the capital.