SRA Logo (print)

Society For Risk Analysis Annual Meeting 2008

Risk Analysis: the Science and the Art

Session Schedule & Abstracts

* Disclaimer: All presentations represent the views of the authors, and not the organizations that support their research. Please apply the standard disclaimer that any opinions, findings, and conclusions or recommendations in abstracts, posters, and presentations at the meeting are those of the authors and do not necessarily reflect the views of any other organization or agency. Meeting attendees and authors should be aware that this disclaimer is intended to apply to all abstracts contained in this document. Authors who wish to emphasize this disclaimer should do so in their presentation or poster. In an effort to make the abstracts as concise as possible and easy for meeting participants to read, the abstracts have been formatted such that they exclude references to papers, affiliations, and/or funding sources. Authors who wish to provide attendees with this information should do so in their presentation or poster.

Common abbreviations

Models, Myths, and Risk-Based Decision Making

Room: Stone   4:00-5:30 PM

Chair(s): William Huber

T4-H.1  16:00  Don't confuse me with the facts! Cain LG*, Hanna C, Linkov I; United States Army Corps of Engineers

Abstract: Consider the theme of this conference; the science and the art. Science is defined as skillful application of facts or principles. Interestingly, art is defined as skillful application of methods or principles. Either can be applied to essentially any human endeavor. Where science is concerned with facts, principles, and methods, art is concerned with principles and methods. Taken together, one may conclude that the arts may not be so concerned with the facts. In practice, risk assessors apply scientific skill to multiple facts, principles, and sources of information. Successful risk assessment requires an instinctive temperament and “artsy” skill. The challenge to bring contrast to the familiar “gray” is only heightened by quips that risk assessors are not encumbered by the facts! This presentation will explore examples of DoD site investigations at military sites with munitions and explosives of concern (MEC) and munition constituents (MC). Quantitative methods to gauge MEC and MC are inherently incompatible with potentially overwhelming human perception regarding physical hazards of explosions or uncompensated loss due to chemical contamination. Multi criteria decision analysis will be discussed as an effective consensus-building decision support tool to successfully join the fray of environmental decision-making. Examples of consensus-based critical information will be provided including the development of transport model partition coefficients for emerging and emergent contaminants, discrepancies in observations versus transport model predictions (calibration), uncertainties in bench scale tests supporting predictive transport models versus actual environmental conditions, and logical use of up-to-the-minute sampling methods like Method 8330B. The opportunities are plentiful for risk assessors to further develop their unique qualities as the decision making “pot” is stirred.

T4-H.2  16:20  Scientific knowledge and Mythology. SERBANESCU D*; PRIVATE PARTICIPATION

Abstract: Acquisition of the human knowledge is a complex process tightly connected with the society and culture. The paper presents a new approach proposed to be used in the evaluation of the paradoxes appearing in the methodologies of various studies and theories (with examples and references from risk analyses) and in the search of solutions and the path forward. The approach is based on the defined isomorphism between scientific knowledge and mythological knowledge and is composed of three steps:Firstly the limits and strengths of the analogy are defined and then a set of proposed rules which are considered to be acting in various phases and for various types of scientific knowledge are formulated in an axiomatic type. Examples are also indicated. Even if they are mainly from the risk and complexity theories (and based on real cases and results in which the author was involved so far), it is considered that the approach is actually applicable (with its nuances of course) to any scientific field.Secondly a set of possible interpretations (based on the isomorphism as defined initially between mythological and scientific knowledge) of the types of paradoxes is defined for every specific phase defined previously for a scientific knowledge process. A group of principles considered to be applicable is formulated and it is defined how they work at each phase and what is their interconnection with paradox solving goal of the endeavor.In final step a set of guiding rules of how to search for solutions when paradoxes appear, which was identified as being typical paradigms for typical type of scientific paradoxes, is presented.

T4-H.3  16:40  Role of participatory approaches in developing risk assessments in equine welfare. Rostom A*, Madany M.M, Childs A.C, Eager R.A; Brooke Hospital for Animals

Abstract: The Brooke is a working equine welfare charity aiming to develop evidence-based interventions tackling priority equine welfare issues. Risk Assessment (RA) methods are used to identify the causes of prioritized welfare issues and provide evidence for use by community and veterinary teams working in the field. Methodology uses participatory tools (PT’s) which enable incorporation of a community’s unique perspective of risk; facilitate data collection in the community and improve acceptance of interventions. PT’s allow members of the animal-owning community to express their knowledge, needs, practices and expectations. Brooke uses PT’s for community inclusion in three areas: 1) Prioritizing welfare issues for RA to investigate. A selection of key stakeholders performs brainstorming exercises and matrix scoring to identify human needs and relate them to animal needs. The resulting recognition of an animal welfare framework has previously been difficult to achieve. Animal welfare then forms the basis of a participatory process in which owners prioritize welfare issues considered most important from both the animal and owner perspective. 2) Sourcing potential risk factors for future measurement. Root cause analysis and focus group discussion are carried out with the community. Knowledge of the community often identifies details of human practices and daily work which may not have otherwise been considered for inclusion in a RA. 3) Social mapping offers a detailed knowledge of local service providers, animal work types and the equine population. Such knowledge ensures a good design of RA and provides a sound basis for implementation. RA tools and methods enable Brooke community and veterinary teams to run intervention and monitoring activities based on community acceptance and well documented knowledge of community dynamics. We believe this level of community participation is vital in the successful running of ‘in field’ RA and the use of RA findings in interventions.

T4-H.4  17:00  Optimal Design of Qualitative Risk Matrices to Classify Quantitative Risks. Huber WA*, Cox LA; Quantitative Decisions; Cox Associates

Abstract: A risk matrix presents the risk associated with paired categories of risk components, such as frequencies and severities of outcomes. As such it discretely approximates a quantitative risk calculation. Relative to a given decision threshold (often interpreted as "acceptable risk" or "risk appetite"), how well can a risk matrix classify risks? It is known that risk matrices are inherently ambiguous, have limited ability to rank quantitative risks correctly, and can result in worse-than-random decisions. Despite these limitations, risk matrices are frequently used for convenience and simplicity. Optimizing their performance can help to limit the harm from using them. This paper shows how to construct optimized risk matrices with respect to either of two criteria: the expected number of misclassification errors or the maximum possible size of misclassification errors. The method has a simple and practical geometric interpretation. The solutions reveal a close connection between minimizing expected number of errors (when risky prospects have a joint uniform distribution of the risk components) and minimizing the maximum possible size of errors for misclassified prospects. It turns out that the same risk matrices can solve both types of optimization problem (for different thresholds, though). In the common case of 5 x 5 risk matrices often used in practice, the optimal designs differ from typical naive designs and can have less than 1/7 their error rate.

T4-H.5  17:20  Self-fulfilling prophecy in health and disease management. A Meta risk. Eisinger F*; Institut Paoli-Calmettes; INSERM UMR 912

Abstract: To be afraid of low risk leads to at least three deleterious side effects. Firstly being afraid is uncomfortable, secondly being afraid of low risks prevents you from mitigating higher risks and lastly precaution could be more hazardous and “costly” than the actual focused risk reduction. One of the counter productive impacts of prevention, a peculiar and ironical one is to increase the risk expected to be avoided by action. Cancer is still defined by a pathologist identifying cancerous cells under microscopic examination. However, moving upstream towards smaller tumors has increased a kind of pathological-clinical dissociation. Indeed, if some of the small tumors identified using early screening should have had evolve towards clinically life-threatening tumors, some of them, if not removed, wouldn’t have progressed. Currently, there is not yet a molecular analysis that could distinct aggressive tumors, which will become life threatening, from the indolent, non-progressing tumors. Therefore, in some cases, early screening is life saving, but in others it is worthless. This has been described as “overdiagnosis” (with its deleterious implication “overtreatment”). Recently it had been observed a paradoxal still higher risk of breast cancer in women who tested negative for a known familial BRCA mutation (an extremely important risk factor for breast cancer). Besides possible epidemiological bias, the authors made the hypothesis of other shared risk factors: genetic or environmental. However, an alternative hypothesis could be made: relatives of women with breast cancer undergo mammography screening more often and at an earlier age than women without such family history. Their mammogram are usually subjected to more cautious interpretation resulting in a higher rate of biopsies; which in turn might increase the rate of diagnosed cancers (only a part of them should have had evolve to a clinical threat). Perceived risk could in this way create a self-fulfilling prophecy.

[back to schedule]