Computational methods for complex stochastic systems: a review of some alternatives to MCMC

We consider analysis of complex stochastic models based upon partial information. MCMC and reversible jump MCMC are often the methods of choice for such problems, but in some situations they can be difficult to implement; and suffer from problems such as poor mixing, and the difficulty of diagnosing convergence. Here we review three alternatives to MCMC methods: importance sampling, the forward-backward algorithm, and sequential Monte Carlo (SMC). We discuss how to design good proposal densities for importance sampling, show some of the range of models for which the forward-backward algorithm can be applied, and show how resampling ideas from SMC can be used to improve the efficiency of the other two methods. We demonstrate these methods on a range of examples, including estimating the transition density of a diffusion and of a discrete-state continuous-time Markov chain; inferring structure in population genetics; and segmenting genetic divergence data.     

Profile upper Confidence Limits from Discrete Data

When the data are discrete, standard approximate confidence limits often have coverage well below nominal for some parameter values. While ad hoc adjustments may largely solve this problem for particular cases, Kabaila & Lloyd (1997) gave a more systematic method of adjustment which leads to tight upper limits, which have coverage which is never below nominal and are as small as possible within a particular class. However, their computation for all but the simplest models is infeasible. This paper suggests modifying tight upper limits by an initial replacement of the unknown nuisance parameter vector by its profile maximum likelihood estimator. While the resulting limits no longer possess the optimal properties of tight limits exactly, the paper presents both numerical and theoretical evidence that the resulting coverage function is close to optimal. Moreover these profile upper limits are much (possibly many orders of magnitude) easier to compute than tight upper limits.     

Evidence of fertility regulation among rural french villagers, 1749–1789

In this first part an econometric model describing the sequences of live births to individual couples is developed and estimated, using data for a sample of married women in rural French communes during the period 1749–1789. The main methodological objective is to devise a means of controlling for unobserved heterogeneities that result in the endogeneity of variables describing a couple's accumulated demographic experience. Without such controls it is not possible to secure unbiased estimates of responses to familial demographic events, such as the births and deaths of children. Such estimates are sought here as indirect indicators of the existence and nature of deliberate fertility control.Un modèle économétrique, décrivant la suite des naissances vivantes d'un couple donné, est développé et estimé dans cette première partie, en utilisant des données sur un échantillon de femmes mariées dans des communes rurales françaises durant la période 1749–1789. Le principal objectif méthodologique est de fournir un moyen de contrôler l'hétérogénéité non observée qui résulte de variables endogènes décrivant l'expérience cumulative d'un couple. Sans un tel contrôle il n'est pas possible d'obtenir des estimations sans biais de l'effet d'événements démographiques, tels que les naissances et les décès d'enfants. De tels estimateurs sont recherchés ici en vue de fournir des indications indirectes sur l'exercice et la nature d'une prévention des naissances délibérée.

---

The second part of this article will be published in theEuropean Journal of Population, Vol. 5, no. 2, pp. 173–206.     

Changing Perceptions through Contact

Contact has long been the major vehicle for changing beliefs and perceptions of minority groups. However, few studies in the area of integration of people with mental retardation have examined underlying assumptions of contact situations. Contact between volunteers and people with mental retardation at a Special Olympics' state games was examined within a theoretical framework wherein mental retardation was viewed as social construct and contact conceptualized within five components of contact theory. Theoretical and empirical study indicated that instances of negative effect limit the extent of change in perceptions the general public hold of people with mental retardation. However, contact, as compared to no contact, does contribute to positive changes in perceptions.     

Do Drivers of Small Cars Take Less Risk in Everyday Driving?

Previously reported observed data on risky everyday driving are brought together and reanalyzed in order to focus on the relation between risky driving and the size of the car being driven, as indicated by car mass. The measures of risky driving include separation between vehicles in heavy freeway traffic and speed on a two lane road. Observed seat belt use provides a third measure of driver risk. Confounding effects arising from the observed association between car mass and driver age are taken into account by segmenting the data into three driver age groups. Driver risk taking is found to increase with increasing car mass for each of these three aspects of everyday driving. The implications of these results with respect to driver fatality rates are discussed in terms of a simple model relating observed risky driving to the likelihood of involvement in a severe crash.     

A sociology of hackers

Illicit computer intruders, or hackers, are often thought of as pathological individuals rather than as members of a community. However, hackers exist within social groups that provide expertise, support, training, journals and conferences. This article outlines this community to establish the nature of hacking within ‘information societies’. To delineate a ‘sociology of hackers’, an introduction is provided to the nature of computer–mediated communication and the act of computer intrusion, the hack. Following this the hacking community is explored in three sections. First, a profile of the number of hackers and hacks is provided by exploring available demographics. Second, an outline of its culture is provided through a discussion of six different aspects of the hacking community. The six aspects are technology, secrecy, anonymity, membership fluidity, male dominance and motivations. Third, an exploration of the community's construction of a boundary, albeit fluid, between itself and its other, the computer security industry, is provided. This boundary is constructed through metaphors whose central role is to establish the ethical nature of hacking. Finally, a conclusion that rejects any pathologisation of hackers is offered.     

Interest Organizations,Information, and Policy Innovation in the U.S. Congress1 2

Interest organizations are hypothesized to strongly affect public policy, but the evidence that they do so is mixed. This article argues that one reason for the disparity is a gap between theory and research: theory suggests that information provided by interest organizations should strongly influence elected officials, but there is no systematic research on its impact. We examine a potentially important source of information for members of the U.S. Congress—testimony at committee hearings—to ascertain if it affects the enactment of policy proposals. The data, based on content analysis of almost 1,000 testimonies on a stratified random sample of policy proposals, describe who testifies, their arguments, and the evidence they provide. Supporters of a proposal emphasize the importance of the problem being addressed, while opponents claim the proposed policy will be ineffective and try to reframe the debate. Information—particularly information regarding policy effectiveness—does affect the likelihood that a policy proposal will be enacted.     

Describing the Dynamics of Attention to TV Commercials: A Hierarchical Bayes Analysis of the Time to Zap an Ad

This paper provides insights into the dynamics of attention to TV commercials via an analysis of the length of time that commercials are viewed before being 'zapped'. The model, which incorporates a flexible baseline hazard rate and captures unobserved heterogeneity across both consumers and commercials using a hierarchical Bayes approach, is estimated on two datasets in which commercial viewing is captured by a passive online device that continually monitors a household's TV viewing. Consistent with previous findings in psychology about the nature of attentional engagement in TV viewing, baseline hazard rates are found to be non-monotonic. In addition, the data show considerable ad-to-ad and household-to-household heterogeneity in zapping behavior. While one of the datasets contains some information on characteristics of the ads, these data do not reveal any firm links between the ad heterogeneity and the ad characteristics. A number of methodological and computational issues arise in the hierarchical Bayes analysis.     

Risk,Media, and Stigma at Rocky Flats

Public responses to nuclear technologies are often strongly negative. Events, such as accidents or evidence of unsafe conditions at nuclear facilities, receive extensive and dramatic coverage by the news media. These news stories affect public perceptions of nuclear risks and the geographic areas near nuclear facilities. One result of these perceptions, avoidance behavior, is a form of "technological stigma" that leads to losses in property values near nuclear facilities. The social amplification of risk is a conceptual framework that attempts to explain how stigma is created through media transmission of information about hazardous places and public perceptions and decisions. This paper examines stigma associated with the U.S. Department of Energy's Rocky Flats facility, a major production plant in the nation's nuclear weapons complex, located near Denver, Colorado. This study, based upon newspaper analyses and a survey of Denver area residents, finds that the social amplification theory provides a reasonable framework for understanding the events and public responses that took place in regard to Rocky Flats during a 6-year period, beginning with an FBI raid of the facility in 1989.