Search

A new paradigm for risk analysis is proposed. It is a holistic conceptualization which recognizes an intrinsic duality between scientific and social risk perspectives. It defines risk states in terms of a minimal but necessary set of properties and qualities of risk issues at various scales of resolution (macro, meso, micro), and aggregations between them. It is founded in a philosophy which recognizes knowledge‐based processes as fundamental both to life (and in turn physical aspects of risk) and to cognition (social aspects of risk). It is through these processes that explanations for risk states are to be found. Implications for risk management are noted.

The process of pollution risk assessment requires the assimilation of data that are spatially variable in nature, making geographical information systems (GIS) an ideal tool for such assessments. Over half of Britain’s drinking water is obtained from surface water abstractions, many of which are situated in upland areas. In order to optimise the quality of abstracted waters it is important to assess the possible risks of pollution upstream of the point of abstraction.

A prototype framework for representing uncertainties in radiological data is introduced. Through this framework, inherent variability in the quality of radiological data can be managed and communicated efficiently, systematically an consistently. Codings derived from the framework have applicability to radiological data, irrespective of the source from which the data are obtained, and irrespective of the context in which they may be used. The coding, in effect, can itself become part of a radiological base.

In this paper, basic concepts of the methodology known as Q -analysis are used to expose aspects of the involvement of local authorities in energy conservation. The involvement of 333 local authorities is examined in respect of 48 different energy conservation activities of potential relevance to them. A great deal of variability is found between different local authorities in terms of the conservation activities they pursue; a variability far exceeding what can be accounted for purely in terms of different statutory functions of different tiers of local government. The Q -analysis brings out this variability quite explicitly and can also be used to suggest where the most useful or plausible modifications in the current pattern of involvement can be sought. Since we invoke only the more elementary aspects of Q -analysis in this paper, we do not explicitly consider notions of so-called traffic -- the degree of involvement by each authority in each of the individual conservation activities, expenditure thereby incurred (or saved), or net energy requirements thereby reduced. This is on the agenda for future work.

Multifactorial data on risk originally collected by Fischoff et al and analysed using the method of factor analysis are reanalysed in this paper using Atkin's Q -analysis algorithm and Johnson's method of Q -discrimination analysis. These algorithms are presented as distinctive methods for interpreting data by revealing their internal structure. Unlike more familiar methods of factor analysis that can be used for a broadly similar purpose, the data are not distorted in any way, for example via the calculation of correlation coefficients. The 'results' which are obtained by invoking Q -analysis are contrasted with those originally obtained from factor analysis, and some comment on the apparent differences is made.

The purpose of this paper is twofold: first to provide a systematic identification of a range of criteria whereby the utility and quality of modelling can be assessed; second to comment on the characteristics of a number of key modelling styles in the light of the criteria distinguished. The goal is to further general understanding of the multiplicity of roles of modelling, and of its various limitations, among interested research and planning communities.

In this paper a recently developed framework designed to facilitate the expression, interpretation, and communication of imperfect quantitative information is adapted in order to elucidate the quality and significance of recently published measures of UK energy efficiency (MONERGY) need and performance. The analysis in the paper uncovers a considerable variability in the quality of the publicised measures; this in turn provides crucial pointers to the way they should be interpreted, and in some cases indicates ground for possible improvement.

The increasing interest that is being shown in uncertainties surrounding the estimation of radiation exposures and consequent health effects calls for the development of new ways of representing the degree of robustness of scientific knowledge in this field. This paper describes part of a new notational system-a pedigree coding-for representing the nature and significance of particular sources of imperfection afflicting the production of quantitative estimates. The pedigree coding is applied to the case of estimating radiation exposure of young people in Seascale as a result of Sellafield discharges. Empirical codings, resulting from interviews with relevant specialists, are shown to provide a cogent way of summarising critical sources of imperfection, this in turn giving crucial clues about the way related quantitative estimates should be communicated and interpreted. Lines for further development and application of the pedigree coding system are suggested.

Some empirical explorations of the use of the Galois lattice approach are presented as an alternative to the more traditional algorithm used by the Q -analysis research community. The explorations are conducted in the context of three fields of study: prey-predator relations, social networks, and clinical psychiatry. It is suggested that the Galois lattice approach is superior to the traditional Q -analysis algorithm on grounds of accuracy, precision, and economy, and that data interpretation previously guided by the traditional algorithm should be reappraised.

A simple conceptual tool for assessing the reliability of calculated model outputs of radiological assessment models is introduced. By its means the different sorts of uncertainties in radiological assessment models can be systematically distinguished, and the combined effects of these different sorts of uncertainty in affecting the reliability of calculated model outputs can be displayed and evaluated. The tool is simple and economical, and also perspicuous and powerful, complementing existing techniques of sensitivity analysis.

Aspects of the publicly acknowledged assessment of risk in siting-decisions in Scotland for potentially hazardous North Sea gas processing-installations are considered. Safety issues were of concern to the respective local authority councils and planning departments, the Health and Safety Executive, the developers seeking planning permission, and, in various instances, port authorities, Secretaries of State, and public-interest groups. The assessment of potential risk to the public by each of these main agencies is analysed and evaluated, illustrating key features from a decade which has seen significant changes in the art of risk assessment in the United Kingdom and elsewhere in the world, and related, though rather more modest, changes in practice.

The incorporation of safety considerations into land-development control decisions associated with nonnuclear hazardous installations in a major metropolitan planning authority area in the United Kingdom is examined. The experience of the case-study authority over the last ten years is examined with reference to the operation of the consultation procedure whereby advice on safety issues is obtained from the Health and Safety Executive, and with reference to the impact that consideration of safety has had on land-use around hazardous sites. Evidence is presented of serious difficulties in planning for nonnuclear major hazard installations and of problems with past and existing consultation procedures. A further aim of the paper is to draw wider conclusions in relation to difficulties in land-use planning around hazardous installations in general. Topical and controversial planning considerations are raised within an empirical context.

In this paper, aspects of four areas of controversy which have recently surrounded the development and application of Q -analysis are examined. It is suggested that the difficulties over which the controversy has been generated are underpinned by important substantive problems, which those wishing to utilise the approach should be aware of, but that further development and application of the style of analysis afforded by Q -analysis and related methods should continue.

In this paper an elementary exploration is given of some of the features of Q-discrimination analysis, an adaptation by Johnson of Atkin's Q-analysis for use in clustering work. The method is summarised. It is shown how the performance of the method varies considerably, depending upon whether the data involved are integer or noninteger, over a given range of values, and on whether they are ordinal or interval. An ambiguity over the criteria for determining discrimination levels is also considered.

A pedagogic exposition of Ho's Galois lattice approach for analysing social data is given in this paper. This provides a basis for respecifying Atkin's Q -analysis in a way that is believed to be more powerful, more economical, and accessible to a wider audience. Concepts of traffic and eccentricity are reexamined, but more generally, the suggested respecification may enhance the utility of Q -analysis and broaden its applicability, while remaining true to its underlying ethos.

A prototype framework for representing uncertainties in radiological model parameters is introduced. This follows an earlier development (Funtowicz et. al. ibid., P.257, 1989) of a corresponding framework for representing uncertainties in radiological data. Refinements and extensions to the earlier framework are needed in order to take account of the additional contextual factors consequent on using data entries to quantify model parameters. The parameter coding can in turn feed in to methods for evaluating uncertainties in calculated model outputs.

This paper establishes the convergence of iterative routines formatrix balancing in which some row and some column totals must equal prespecified positive numbers. This balancing procedure is called a modified biproportional problem. Applications of the procedure are cited, various theoretical properties discussed, and its relationship with the full biproportional problem, treated in an earlier paper (Macgill, 1977a), is noted.