Uncommon Priors Require Origin Disputes

In standard belief models, priors are always common knowledge. This prevents such models from representing agents’ probabilistic beliefs about the origins of their priors. By embedding standard models in a larger standard model, however, pre-priors can describe such beliefs. When an agent’s prior and pre-prior are mutually consistent, he must believe that his prior would only have been different in situations where relevant event chances were different, but that variations in other agents’ priors are otherwise completely unrelated to which events are how likely. Due to this, Bayesians who agree enough about the origins of their priors must have the same priors.     

Proxies in the New Political Economy: Caveat Emptor

Some common professional criteria for evaluating econometric tests that use proxies for political, social, and institutional variables as regressors may lull scholars into a false sense of security about specious results. The highly publicized correlations between income and economic freedom based on the economic freedom indexes supplied by the Fraser Institute, Heritage Foundation, and Freedom House are used as illustrations. Because these correlations bear strong resemblances to one another, most economists would consider them mutually supportive. Yet this is not so, implying that mutual resemblance and consistency are not criteria for judging econometric results. (JEL B49 , C19 )     

NEWS ABOUT MURDER IN AN AFRICAN AMERICAN NEWSPAPER: Effects of Relative Frequency and Race and Gender Typifications

Previous scholarly research on selection bias in news about murder indicates that race and gender stereotypes and, to a much lesser extent, the relative frequency of particular murders explain why some homicides are made into news and others are not. However, previous research has directed nearly exclusive attention to white newspapers. The present research remedies this omission by directing attention to the factors that shape selection bias in news about murder in a big-city African American newspaper. The results indicate that the relative frequency dimension of newsworthiness is a weak and inconsistent explanation of selection bias in news about murder in the black newspaper examined. Race, however, has profoundly different effects in white and African American newspapers. Whereas white newspapers use long-standing race stereotypes to filter news about murder, the data from the African American newspaper signal a firm rejection of the black image in the white mind (Entman and Rojecki 2000). Newspaper images of women and men are another matter. In both black and white newspapers, gender stereotypes uniformly filter news about murder and fundamentally distort gender effects. The implications of these findings are discussed including the clear need and ample scholarly room for replicative analyses of news about murder in other African American newspapers, with a keen eye on both "raced ways of seeing" (Hunt 1999, pp. 181–215) and gendered ways of seeing.     

Working Beyond 65: Predictors of Late Retirement for Women and Men MBAs

Comparisons of those who planned to continue working after the age of 65 with those who did not were made for 145 women and 414 men working in managerial fields. All received MBA degrees between the years of 1973 and 1982. About 20% definitely wanted to work after age 65. More positive views of work were predictive of wanting to continue working as was having nontraditional gender-role attitudes. Men who planned to continue working were particularly likely to have a spouse wanting to work past the age of 65. Several other factors appeared to operate differently for women and men.     

Video Deficit in Toddlers’ Object Retrieval: What Eye Movements Reveal About Online Cognition

The purpose of this study was to determine whether toddlers exhibit different eye‐movement patterns when watching real events versus video demonstrations in an object‐retrieval task. Twenty‐four‐month‐olds (N = 36) searched for a sticker on a felt board after watching an experimenter hide it behind a felt object in person or via video. Eye movements during the hiding event were recorded. Compared to those watching in‐person events, children watching video spent more time looking at the target location overall, yet they had relatively poor search performance. Visual attention to the target location predicted search performance in the video condition only; children who watched in‐person hiding events had high success rates even if they paid relatively little visual attention to the correct location. Findings are consistent with the hypothesis that toddlers process information more quickly for in‐person (versus video) events, enabling them to learn as well (or better) despite relatively low selective attention. Thus, relatively poor encoding, as well as memory retrieval, may underlie the video deficit.     

Modelling stochastic order in the analysis of receiver operating characteristic data: Bayesian non-parametric approaches

Summary.  The evaluation of the performance of a continuous diagnostic measure is a commonly encountered task in medical research. We develop Bayesian non-parametric models that use Dirichlet process mixtures and mixtures of Polya trees for the analysis of continuous serologic data. The modelling approach differs from traditional approaches to the analysis of receiver operating characteristic curve data in that it incorporates a stochastic ordering constraint for the distributions of serologic values for the infected and non-infected populations. Biologically such a constraint is virtually always feasible because serologic values from infected individuals tend to be higher than those for non-infected individuals. The models proposed provide data-driven inferences for the infected and non-infected population distributions, and for the receiver operating characteristic curve and corresponding area under the curve. We illustrate and compare the predictive performance of the Dirichlet process mixture and mixture of Polya trees approaches by using serologic data for Johne's disease in dairy cattle.     

Bayesian non-parametric models for regional prevalence estimation

We developed a flexible non-parametric Bayesian model for regional disease-prevalence estimation based on cross-sectional data that are obtained from several subpopulations or clusters such as villages, cities, or herds. The subpopulation prevalences are modeled with a mixture distribution that allows for zero prevalence. The distribution of prevalences among diseased subpopulations is modeled as a mixture of finite Polya trees. Inferences can be obtained for (1) the proportion of diseased subpopulations in a region, (2) the distribution of regional prevalences, (3) the mean and median prevalence in the region, (4) the prevalence of any sampled subpopulation, and (5) predictive distributions of prevalences for regional subpopulations not included in the study, including the predictive probability of zero prevalence. We focus on prevalence estimation using data from a single diagnostic test, but we also briefly discuss the scenario where two conditionally dependent (or independent) diagnostic tests are used. Simulated data demonstrate the utility of our non-parametric model over parametric analysis. An example involving brucellosis in cattle is presented.