BIAS IN LIST-ASSISTED TELEPHONE SAMPLES

A number of researchers have suggested list-assisted samplingfor the selection of telephone households to overcome some ofthe operational difficulties associated with the Mitofsky-Waksbergmethods of random digit dialing (RDD). An advantage of a list-assistedmethod of RDD is that an equal probability systematic sampleof telephone numbers can be selected and the variances of estimatesfrom such a sample are usually lower than from a clustered designlike the Mitofsky-Waksberg method. The main disadvantage ofthe list-assisted method is that it excludes some householdsfrom the sample, thus creating a coverage bias in the estimates.This article describes research on the coverage bias for a particularmethod of list-assisted sampling. The two key determinants ofcoverage bias are the proportion of households that are noteligible for the sample and the differences in the characteristicsof the covered and not covered populations. The results showthat about 4 percent of all households are excluded in nationalsamples using this method of sampling. Furthermore, they showthat the differences between the covered and uncovered populationsare generally not large. The coverage bias resulting from theseconditions may often be small.     

Ethnicity, migration and health: A population-based study of 338 refugees from Latin America, 396 nonrefugee immigrants from Finland and 161 from southern Europe and 996 age-, sex- and education-matched controls from Sweden

This article shows the influence of being a refugee from Latin America or a nonrefugee immigrant from southern Europe or Finland on self-reported illness, controlling for social factors and lifestyle. The study population consisted of 338 Latin American refugees, a random sample of 396 Finnish and 161 southern European immigrants and 996 age-, sex- and education-matched Swedish controls. The data were analysed unmatched with logistic regression (multivariate analysis) in main effect models. The strongest independent risk indicator for long-term illness was being a Latin American refugee (estimated odds ratio (OR)=2.96, 95% confidence interval (CI)=2.19–3.82). There was a significant association between being a Latin American refugee and period prevalence, ill health and unsatisfied need for care. Being a southern European or Finnish immigrant was a risk indicator of ill health but was not associated with the other dependent factors. Not feeling secure in daily life was a strong risk indicator for long-term illness and ill health (estimated OR=1.89, 95% CI=1.26–2.76 and OR=3.04, 95% CI= 1.97–4.48) respectively). Being a Latin American refugee was equal in importance to traditional risk factors such as overweight and not taking regular exercise for long-term illness and ill health.     

The arithmetic of growth: Methods of calculation

This is a tutorial on the relations between population data and the rates of growth that are calculated from the data. For the calculation of rates of growth, discrete and continuous compounding will be compared so that the reader can see the reasons for using the mathematics of continuous compounding, which is the mathematics of exponential growth. Some properties of exponential growth are developed. Semi-logarithmic graphs will be discussed as a device for representing the size of growing populations and for analyzing the nature of the growth. Illustrative examples will be worked out in order to emphasize applications and utility.     

Multiparameter evidence synthesis in epidemiology and medical decision-making: current approaches

Summary.  Alongside the development of meta-analysis as a tool for summarizing research literature, there is renewed interest in broader forms of quantitative synthesis that are aimed at combining evidence from different study designs or evidence on multiple parameters. These have been proposed under various headings: the confidence profile method, cross-design synthesis, hierarchical models and generalized evidence synthesis. Models that are used in health technology assessment are also referred to as representing a synthesis of evidence in a mathematical structure. Here we review alternative approaches to statistical evidence synthesis, and their implications for epidemiology and medical decision-making. The methods include hierarchical models, models informed by evidence on different functions of several parameters and models incorporating both of these features. The need to check for consistency of evidence when using these powerful methods is emphasized. We develop a rationale for evidence synthesis that is based on Bayesian decision modelling and expected value of information theory, which stresses not only the need for a lack of bias in estimates of treatment effects but also a lack of bias in assessments of uncertainty. The increasing reliance of governmental bodies like the UK National Institute for Clinical Excellence on complex evidence synthesis in decision modelling is discussed.     

Spatial variation in soil inorganic nitrogen across an arid urban ecosystem

We explored variations in inorganic soil nitrogen (N) concentrations across metropolitan Phoenix, Arizona, and the surrounding desert using a probability-based synoptic survey. Data were examined using spatial statistics on the entire region, as well as for the desert and urban sites separately. Concentrations of both NO₃-N and NH₄-N were markedly higher and more heterogeneous amongst urban compared to desert soils. Regional variation in soil NO₃-N concentration was best explained by latitude, land use history, population density, along with percent cover of impervious surfaces and lawn, whereas soil NH₄-N concentrations were related to only latitude and population density. Within the urban area, patterns in both soil NO₃-N and NH₄-N were best predicted by elevation, population density and type of irrigation in the surrounding neighborhood. Spatial autocorrelation of soil NO₃-N concentrations explained 49% of variation among desert sites but was absent between urban sites. We suggest that inorganic soil N concentrations are controlled by a number of ‘local’ or ‘neighborhood’ human-related drivers in the city, rather than factors related to an urban-rural gradient.     

Computing maximum likelihood estimates from type II doubly censored exponential data

It is well-known that, under Type II double censoring, the maximum likelihood (ML) estimators of the location and scale parameters, θ and δ, of a twoparameter exponential distribution are linear functions of the order statistics. In contrast, when θ is known, theML estimator of δ does not admit a closed form expression. It is shown, however, that theML estimator of the scale parameter exists and is unique. Moreover, it has good large-sample properties. In addition, sharp lower and upper bounds for this estimator are provided, which can serve as starting points for iterative interpolation methods such as regula falsi. Explicit expressions for the expected Fisher information and Cramér-Rao lower bound are also derived. In the Bayesian context, assuming an inverted gamma prior on δ, the uniqueness, boundedness and asymptotics of the highest posterior density estimator of δ can be deduced in a similar way. Finally, an illustrative example is included.     

Bayesian modelling of catch in a north-west Atlantic fishery

Summary. We model daily catches of fishing boats in the Grand Bank fishing grounds. We use data on catches per species for a number of vessels collected by the European Union in the context of the Northwest Atlantic Fisheries Organization. Many variables can be thought to influence the amount caught: a number of ship characteristics (such as the size of the ship, the fishing technique used and the mesh size of the nets) are obvious candidates, but one can also consider the season or the actual location of the catch. Our database leads to 28 possible regressors (arising from six continuous variables and four categorical variables, whose 22 levels are treated separately), resulting in a set of 177 million possible linear regression models for the log-catch. Zero observations are modelled separately through a probit model. Inference is based on Bayesian model averaging, using a Markov chain Monte Carlo approach. Particular attention is paid to the prediction of catches for single and aggregated ships.