A Nonstationary Negative Binomial Time Series with Time-Dependent Covariates: Enterococcus Counts in Boston Harbor

Boston Harbor has had a history of poor water quality, including contamination by enteric pathogens. We conduct a statistical analysis of data collected by the Massachusetts Water Resources Authority (MWRA) between 1996 and 2002 to evaluate the effects of court-mandated improvements in sewage treatment. Motivated by the ineffectiveness of standard Poisson mixture models and their zero-inflated counterparts, we propose a new negative binomial model for time series of Enterococcus counts in Boston Harbor, where nonstationarity and autocorrelation are modeled using a nonparametric smooth function of time in the predictor. Without further restrictions, this function is not identifiable in the presence of time-dependent covariates; consequently we use a basis orthogonal to the space spanned by the covariates and use penalized quasi-likelihood (PQL) for estimation. We conclude that Enterococcus counts were greatly reduced near the Nut Island Treatment Plant (NITP) outfalls following the transfer of wastewaters from NITP to the Deer Island Treatment Plant (DITP) and that the transfer of wastewaters from Boston Harbor to the offshore diffusers in Massachusetts Bay reduced the Enterococcus counts near the DITP outfalls.

A Diagnostic Test for the Mixing Distribution in a Generalised Linear Mixed Model

We introduce a diagnostic test for the mixing distribution in a generalised linear mixed model. The test is based on the difference between the marginal maximum likelihood and conditional maximum likelihood estimates of a subset of the fixed effects in the model. We derive the asymptotic variance of this difference, and propose a test statistic that has a limiting chi-square distribution under the null hypothesis that the mixing distribution is correctly specified. For the important special case of the logistic regression model with random intercepts, we evaluate via simulation the power of the test in finite samples under several alternative distributional forms for the mixing distribution. We illustrate the method by applying it to data from a clinical trial investigating the effects of hormonal contraceptives in women.

A unified approach to modeling multivariate binary data using copulas over partitions

Many seemingly disparate approaches for marginal modeling have been developed in recent years. We demonstrate that many current approaches for marginal modeling of correlated binary outcomes produce likelihoods that are equivalent to the proposed copula-based models herein. These general copula models of underlying latent threshold random variables yield likelihood based models for marginal fixed effects estimation and interpretation in the analysis of correlated binary data. Moreover, we propose a nomenclature and set of model relationships that substantially elucidates the complex area of marginalized models for binary data. A diverse collection of didactic mathematical and numerical examples are given to illustrate concepts.

Multilevel Latent Class Models with Dirichlet Mixing Distribution

Latent class analysis (LCA) and latent class regression (LCR) are widely used for modeling multivariate categorical outcomes in social sciences and biomedical studies. Standard analyses assume data of different respondents to be mutually independent, excluding application of the methods to familial and other designs in which participants are clustered. In this paper, we develop multilevel latent class model, in which subpopulation mixing probabilities are treated as random effects that vary among clusters according to a common Dirichlet distribution. We apply the Expectation-Maximization (EM) algorithm for model fitting by maximum likelihood (ML). This approach works well, but is computationally intensive when either the number of classes or the cluster size is large. We propose a maximum pairwise likelihood (MPL) approach via a modified EM algorithm for this case. We also show that a simple latent class analysis, combined with robust standard errors, provides another consistent, robust, but less efficient inferential procedure. Simulation studies suggest that the three methods work well in finite samples, and that the MPL estimates often enjoy comparable precision as the ML estimates. We apply our methods to the analysis of comorbid symptoms in the Obsessive Compulsive Disorder study. Our models' random effects structure has more straightforward interpretation than those of competing methods, thus should usefully augment tools available for latent class analysis of multilevel data.

Copper and human health: biochemistry, genetics, and strategies for modeling dose-response relationships

Copper (Cu) and its alloys are used extensively in domestic and industrial applications. Cu is also an essential element in mammalian nutrition. Since both copper deficiency and copper excess produce adverse health effects, the dose-response curve is U-shaped, although the precise form has not yet been well characterized. Many animal and human studies were conducted on copper to provide a rich database from which data suitable for modeling the dose-response relationship for copper may be extracted. Possible dose-response modeling strategies are considered in this review, including those based on the benchmark dose and categorical regression. The usefulness of biologically based dose-response modeling techniques in understanding copper toxicity was difficult to assess at this time since the mechanisms underlying copper-induced toxicity have yet to be fully elucidated. A dose-response modeling strategy for copper toxicity was proposed associated with both deficiency and excess. This modeling strategy was applied to multiple studies of copper-induced toxicity, standardized with respect to severity of adverse health outcomes and selected on the basis of criteria reflecting the quality and relevance of individual studies. The use of a comprehensive database on copper-induced toxicity is essential for dose-response modeling since there is insufficient information in any single study to adequately characterize copper dose-response relationships. The dose-response modeling strategy envisioned here is designed to determine whether the existing toxicity data for copper excess or deficiency may be effectively utilized in defining the limits of the homeostatic range in humans and other species. By considering alternative techniques for determining a point of departure and low-dose extrapolation (including categorical regression, the benchmark dose, and identification of observed no-effect levels) this strategy will identify which techniques are most suitable for this purpose. This analysis also serves to identify areas in which additional data are needed to better define the characteristics of dose-response relationships for copper-induced toxicity in relation to excess or deficiency.

Review of assessment, design, and mitigation of multiple hazards

Large parts of the world are subjected to one or more natural hazards, such as earthquakes, tsunamis, landslides, tropical storms (hurricanes, cyclones and typhoons), costal inundation and flooding. Virtually the entire world is at risk of man-made hazards. In recent decades, rapid population growth and economic development in hazard-prone areas have greatly increased the potential of multiple hazards to cause damage and destruction of buildings, bridges, power plants, and other infrastructure; thus posing a grave danger to the community and disruption of economic and societal activities. Although an individual hazard is significant in many parts of the United States (U.S.), in certain areas more than one hazard may pose a threat to the constructed environment. In such areas, structural design and construction practices should address multiple hazards in an integrated manner to achieve structural performance that is consistent with owner expectations and general societal objectives. The growing interest and importance of multiple-hazard engineering has been recognized recently. This has spurred the evolution of multiple-hazard risk-assessment frameworks and development of design approaches which have paved way for future research towards sustainable construction of new and improved structures and retrofitting of the existing structures. This report provides a review of literature and the current state of practice for assessment, design and mitigation of the impact of multiple hazards on structural infrastructure. It also presents an overview of future research needs related to multiple-hazard performance of constructed facilities.

La Yeguada volcanic complex, western Panama : an assessment of the geologic hazards using new 40AR/39Ar Ages

La Yeguada volcanic complex is one of three Quaternary volcanic centers in Panama, and is located on the southern slope of the Cordillera Central mountain range in western Panama, province of Veraguas. To assess potential geologic hazards, this study focused on the main dome complex near the village of La Laguna and also examined a cinder cone 10 km to the northwest next to the village of Media Luna. Based on newly obtained 40Ar/39Ar ages, the most recent eruption occurred approximately 32 000 years ago at the Media Luna cinder cone, while the youngest dated eruption at the main dome complex occurred 0.357 ± 0.019 Ma, producing the Castillo dome unit. Cerro Picacho is a separate dome located 1.5 km east of the main complex with a date of 4.47 ± 0.23 Ma, and the El Satro Pyroclastic Flow unit surrounds the northern portion of the volcanic complex and has an age of 11.26 ± 0.17 Ma. No Holocene (10 000 years ago to present) activity is recorded at the La Yeguada volcanic complex and therefore, it is unlikely to produce another eruption. The emergence of a new cinder cone is a possibility, but the associated hazards tend to be low and localized, and this does not pose a significant threat to the small communities scattered throughout the area. The main geologic hazard at the La Yeguada volcanic complex is from landslides coming off the many steep slopes.

Assessment of soil erosion vulnerability in rural communities in the vicinity of Motozintla de Mendoza, Chiapas, Mexico : a physical and sociological approach

Soil erosion is a natural geological phenomenon resulting from removal and transportation of soil particles by water, wind, ice and gravity. As soil erosion may be affected from cultural factors as well. The physical and social phenomena of soil erosion are researched in six communities in the upper part of Rio Grijalva Basin in the vicinity of Motozintla de Mendoza, Chiapas, Mexico. For this study, the USDA RUSLE model was applied to estimate soil erosion rates in the six communities based on the available data. The RUSLE model is based on soil properties, topography, and land cover and management factors. These results showed that estimated soil erosion rates ranged from a high of 2,050 metric ton ha-1 yr-1 to a low of 100 metric ton ha-1 yr-1. A survey concerning knowledge, attitudes and practices (KAP) related to soil erosion was also conducted in all 236 households in the six communities. The main findings of the KAP survey were: 69% of respondents did not know what soil erosion was, while over 40% of the population perceived that hurricanes are the biggest factors that cause soil erosion, and about 20 % of the interviewees said that the landslides are the consequences of the soil erosion. People in communities did not perceive cultural factors as important in conservation efforts for reduce vulnerability to erosion; therefore, the results obtained are suggested to be useful for informing efforts to educate stakeholders.