Book Reviews

Books reviewed:
Curtis Travis, World Risk Society
D. Warner North, Science at EPA: Information in the Regulatory Process     

Variability in Physical Constant Parameter Values from Standard Data Sources and the Implication of this Variability for Risk Analysis

Physical property values are used in environmental risk assessments to estimate media and risk-based concentrations. Recently, however, considerable variability has been reported with such values. To evaluate potential variability in physical parameter values supporting a variety of regulatory programs, eight data sources were chosen for evaluation, and chemicals appearing in at least four sources were selected. There were 755 chemicals chosen. In addition, chemicals in seven environmentally important subgroups were also identified for evaluation. Nine parameters were selected for analysis--molecular weight (MolWt), melting point (MeltPt), boiling point (BoilPt), vapor pressure (VP), water solubility (AqSOL), Henry's law constant (HLC), octanol-water partition coefficient (Kow), and diffusion coefficients in air (Dair) and water (Dwater). Results show that while 71% of constituents had equal MolWts across data sources, <3% of the constituents had equivalent parameter values across data sources for AqSOL, VP, or HLC. Considerable dissimilarity between certain sources was also observed. Furthermore, measures of dispersion showed considerable variation in data sets for Kow, VP, AqSOL, and HLC compared to measures for MolWt, MeltPt, BoilPt, or Dwater. The magnitude of the observed variability was also noteworthy. For example, the 95th percentile ratio of maximum/minimum parameter values ranged from 1.0 for MolWt to well over 1.0 x 10(6) for VP and HLC. Risk and exposure metrics also varied by similar magnitudes. Results with environmentally important subgroups were similar. These results show that there is considerable variability in physical parameter values from standard sources, and that the observed variability could affect potential risk estimates and perhaps risk management decisions.     

Modelling and estimating heavy-tailed non-homogeneous correlated queues: Pareto-inverse gamma HGLM with covariates

Evidence of communication traffic complexity reveals correlation in a within-queue and heterogeneity among queues. We show how a random-effect model can be used to accommodate these kinds of phenomena. We apply a Pareto distribution for arrival (service) time of individual queue for given arrival (service) rate. For modelling potential correlation in arrival (service) times within a queue and heterogeneity of the arrival (service) rates among queues, we use an inverse gamma distribution. This modelling approach is then applied to the cache access log data processed through an Internet server. We believe that our approach is potentially useful in the area of network resource management.     

Does EPA Underestimate Cancer Risks by Ignoring Susceptibility Differences?

A 2009 report of the National Research Council (NRC) recommended that the U.S. Environmental Protection Agency (EPA) increase its estimates of increased cancer risk from exposure to environmental agents by ～7‐fold, due to an approximate ～25‐fold typical ratio between the median and upper 95th percentile persons’ cancer sensitivity assuming approximately lognormally distributed sensitivities. EPA inaction on this issue has raised concerns that cancer risks to environmentally exposed populations remain systematically underestimated. This concern is unwarranted, however, because EPA point estimates of cancer risk have always pertained to the average, not the median, person in each modeled exposure group. Nevertheless, EPA has yet to explain clearly how its risk characterization and risk management policies concerning individual risks from environmental chemical carcinogens do appropriately address broad variability in human cancer susceptibility that has been a focus of two major NRC reports to EPA concerning its risk assessment methods.     

Risk Compensation and Bicycle Helmets

This study investigated risk compensation by cyclists in response to bicycle helmet wearing by observing changes in cycling behavior, reported experience of risk, and a possible objective measure of experienced risk. The suitability of heart rate variability (HRV) as an objective measure of experienced risk was assessed beforehand by recording HRV measures in nine participants watching a thriller film. We observed a significant decrease in HRV in line with expected increases in psychological challenge presented by the film. HRV was then used along with cycling pace and self‐reported risk in a field experiment in which 35 cyclist volunteers cycled 0.4 km downhill, once with and once without a helmet. Routine helmet users reported higher experienced risk and cycled slower when they did not wear their helmet in the experiment than when they did wear their helmet, although there was no corresponding change in HRV. For cyclists not accustomed to helmets, there were no changes in speed, perceived risk, or any other measures when cycling with versus without a helmet. The findings are consistent with the notion that those who use helmets routinely perceive reduced risk when wearing a helmet, and compensate by cycling faster. They thus give some support to those urging caution in the use of helmet laws.     

汉语常用词的稳定性与变异性

以五类常用词——即吃类词、打类词、看类词、听类词、走类词为研究对象,对其进行了历时演变与共时分布的考察,认为汉语常用词既有稳定性,也有变异性,变异是主要的。影响常用词变异的因素主要有社会变化、方言地位的升降、词汇系统的内部机制以及口语与书面语系统的相互交换等。     

电子商务网站动态模糊综合评价模型及实例

电子商务网站评价是衡量网站价值、分析网站优劣、指导网站建设的重要依据和方法。现有评价模型忽视了评价过程中的动态性、模糊性因素，存在很多缺点。动态模糊综合评价法基于动态模糊理论，模拟评价指标的动态性和模糊性，采用多级动态模糊递归方式，减少人为因素的影响。评价结果可以反映评价目标的动态发展趋势，且可以指导网站质量改进。本文最后通过对某商务网站的综合评价验证了该方法的有效性和实用性。     

A new variability order based on tail-heaviness

In this paper, we introduce a new variability order that can be interpreted in terms of tail-heaviness which we will call the tail dispersive order. We provide the new definition, its interpretation and properties and the main characterization. We also study the relationship with other classical variability orders. Finally, we study the tail dispersive order in some classical parametric families and provide some applications in insurance and finance. We conclude with a numerical example applied to log returns distributions.     

Estimating Drug Shelf-life with Unknown Lot-to-lot Variability

This article addresses the problem of estimating the shelf-life of a drug through a standard stability study that is routinely performed in the pharmaceutical industry. Based upon a dataset of the degradation of the drug efficacy over time, the objective is to estimate at what time point the drug efficacy falls below a minimally acceptable level. In this article, attention is directed toward the different roles played by lot-to-lot variability and measurement error in the performances of various estimation methodologies. The standard estimation method, as proposed by the United States Food and Drug Administration, for example, in accordance with current international guidelines, is compared with some new approaches. Simulation results are presented, which illustrate the advantages and disadvantages of the different estimation methodologies, and some recommendations and suggestions are made for shelf-life estimations.