Designing and integrating composite networks for monitoring multivariate gaussian pollution fields

Networks of ambient monitoring stations are used to monitor environmental pollution fields such as those for acid rain and air pollution. Such stations provide regular measurements of pollutant concentrations. The networks are established for a variety of purposes at various times so often several stations measuring different subsets of pollutant concentrations can be found in compact geographical regions. The problem of statistically combining these disparate information sources into a single 'network' then arises. Capitalizing on the efficiencies so achieved can then lead to the secondary problem of extending this network. The subject of this paper is a set of 31 air pollution monitoring stations in southern Ontario. Each of these regularly measures a particular subset of ionic sulphate, sulphite, nitrite and ozone. However, this subset varies from station to station. For example only two stations measure all four. Some measure just one. We describe a Bayesian framework for integrating the measurements of these stations to yield a spatial predictive distribution for unmonitored sites and unmeasured concentrations at existing stations. Furthermore we show how this network can be extended by using an entropy maximization criterion. The methods assume that the multivariate response field being measured has a joint Gaussian distribution conditional on its mean and covariance function. A conjugate prior is used for these parameters, some of its hyperparameters being fitted empirically.     

Factors influencing Soay sheep survival

We present a survival analysis of Soay sheep mark recapture and recovery data. Unlike previous conditional analyses, it is not necessary to assume equality of recovery and recapture probabilities; instead these are estimated by maximum likelihood. Male and female sheep are treated separately, with the higher numbers and survival probabilities of the females resulting in a more complex model than that used for the males. In both cases, however, age and time aspects need to be included and there is a strong indication of a reduction in survival for sheep aged 7 years or more. Time variation in survival is related to the size of the population and selected weather variables, by using logistic regression. The size of the population significantly affects the survival probabilities of male and female lambs, and of female sheep aged 7 or more years. March rainfall and a measure of the North Atlantic oscillation are found to influence survival significantly for all age groups considered, for both males and females. Either of these weather variables can be used in a model. Several phenotypic and genotypic individual covariates are also fitted. The only covariate which is found to influence survival significantly is the type of horn of first-year female sheep. There is a substantial variation in the recovery probabilities over time, reflecting in part the increased effort when a population crash was expected. The goodness of fit of the model is checked by using graphical procedures.     

Bayes linear analysis for graphical models: The geometric approach to local computation and interpretive graphics

This paper concerns the geometric treatment of graphical models using Bayes linear methods. We introduce Bayes linear separation as a second order generalised conditional independence relation, and Bayes linear graphical models are constructed using this property. A system of interpretive and diagnostic shadings are given, which summarise the analysis over the associated moral graph. Principles of local computation are outlined for the graphical models, and an algorithm for implementing such computation over the junction tree is described. The approach is illustrated with two examples. The first concerns sales forecasting using a multivariate dynamic linear model. The second concerns inference for the error variance matrices of the model for sales, and illustrates the generality of our geometric approach by treating the matrices directly as random objects. The examples are implemented using a freely available set of object-oriented programming tools for Bayes linear local computation and graphical diagnostic display.     

Birth outcome, not pregnancy process: Reply to van der Veen

In a recent article (Frisbie. Forbes. and Pullum 1996) we documented racial/ethnic differences in birth outcomes according to a more fine-grained classification than has typically been employed in the demographic literature. In his commentary, van der Veen focuses on the measurement of one of the dimensions of that classification, maturity of the infant. as proxied by the fetal growth ratio. The crux of the critique is easily seen in van der Veen's statement that “all of my disagreements with Frisbie et al. 's method arise from their particular use of a postnatal standard for the assessment of intrauterine growth.” Our critic misunderstands our objective: He fails to realize our interest in birth outcome, not pregnancy process, and does not perceive that our intent was to extend the research extant in both the demographic and public-health literatures in which patently postnatal (i.e., ex utero) measures are taken as outcomes interesting in their own right and/or as risk factors for infant mortality and infant and childhood morbidity. Specifically, he does recognize that we purposefully expanded our focus to include moderately compromised births to determine if they were at higher risk than the normal births with whom they are conventionally categorized. Our discussion draws on research cited in the original article, on studies cited by our critic, and on a few more recent investigations. Although we have never argued that ours is the only, or even the best, approach in all cases, we try to clarify the rationale for, and adduce additional empirical evidence of, the utility of the method we used.     

Drug dependency and HIV testing among state prisoners

HIV and drug use are higher among prisoners than the general US population. This study examines drug dependency/use and differences between prisoners who volunteered for HIV testing and those who did not in a less densely populated state. It was hypothesized that prisoners who volunteered for an HIV test were engaged in more drug use and other risky behaviors than those who did not. Survey data were collected from 600 randomly selected inmates (567 males and 33 females) from 15 state prisons. Subjects were male (95%), white (63%), never married (43%), and 44% volunteered for an HIV test since entering prison. Ninety-two percent of inmates met DSM criteria for drug dependence in their lifetime. Those who volunteered for HIV testing were 2.6 times more likely to ever have used PCP; 1.5 times more likely to ever have used cocaine; 1.4 times more likely to ever have had a problem with drugs; 1.3 times more likely to have used opiates, and 1.6 times more likely to report having been sexually or physically abused. Implications for interventions are discussed.     

An Investigation of Measures of Modern and Old-Fashioned Sexism

Working from the Principle of Beneficence we make the case that there is a need for adequate measures of sexism in men and women. Swim, Aikin, Hall and Hunter (1995) suggest that modern sexism represents a constellation of beliefs in which individuals: (a) disavow women's present-day discrimination; (b) reject women's demands for political and economic power; and (c) disapprove of policies designed to promote gender equality. To assess this construct, Swim et al. (1995) developed the Modern Sexism Scale (MSS), and to date, few studies have assessed its reliability and factor structure. Using data from 187 Canadian participants, confirmatory factor analysis was used to assess the unidimensionality of the MSS and Swim and associates' (1995) Old-Fashioned Sexism Scale (OFSS). Results do not support Swim et al.'s conceptualization of modern sexism as a unidimensional construct; however, the OFSS fit a unidimensional model. The implications of these findings are discussed.     

Communicating future water needs to an at-risk population: lessons learned following defeat of the Applewhite Dam and Reservoir Project in San Antonio,Texas

San Antonio, Texas, the seventh largest city in the United States, has experienced steady population growth, since the “boom” of the 1960s. Projected water shortages due to this growth were realized as early as the 1970s by city leaders and south-central Texas regional development decision makers. To reduce dependence on the already over-taxed, Edwards aquifer, a solution, the Applewhite Dam and Reservoir Project, was developed with wide acceptance by federal, state, and city leaders who regarded the project as a necessary measure for regional growth and development. However, opposition by taxpayer and environmental groups led to referendums of 1991 and 1994 in which voters blocked construction of the dam and reservoir leaving the city with limited options for water provision. This case study investigated the factors which led to a clear mismatch in communication between decision makers—those who were aware of the actual and quantifiable risk to the region in terms of reduced water supplies—and the general public, a population that did not have complete and/or adequate knowledge of their actual risk regarding future water shortages, nor, of solutions being developed, such as the Applewhite project. The findings from this case study indicate that when municipal leadership fails to adequately communicate risk regarding resource shortage to an affected public, as well as, openly planned solutions, that voters are likely to underestimate future impacts of water shortages, heed last-minute opposition, and reject long-standing, publicly proposed projects. The intent of this research is not to support either side in the Applewhite controversy, but to shed perspective on the process of adequately and effectively communicating future water needs to an at-risk population. Decision makers in cities across the United States who are faced with solving problems of limited resources needed by a large populace may be informed by the results of this research.