Society For Risk Analysis Annual Meeting 2017

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

Foundational Issues in Risk Analysis III

Room: Salon J   10:30 am–12:00 pm

Chair(s): Seth Guikema

Sponsored by Foundational Issues in Risk Analysis Specialty Group

W2-H.1  10:30 am  Core Subjects and Principles of Risk Analysis. Aven T*; University of Stavanger, Norway

Abstract: In a recent document from the Society for Risk Analysis a list of core subjects of risk analysis is presented ( It captures five main categories of subjects: Fundamentals (science, knowledge, uncertainties, risk - other basic concepts); Risk assessment; Risk perception and communication; Risk management and governance, and Solving real risk problems and issues. In the talk I will discuss the background, rationale, purpose and importance of this document. I will also reflect on a related issue; what are the key fundamental principles ensuring prudent risk analysis? There is a strong need for authoritative guidance on how to best conduct risk analysis but is it really possible to formulate meaningful and useful principles on a generic level?

W2-H.2  10:50 am  Quantitative Risk Modeling and Management of Interdependent Complex Systems of Systems. Haimes YY*; University of Virginia

Abstract: In this presentation, we focus on characterization and modeling of complex SoS (“Complex SoS”); namely, on their interconnectedness and interdependencies (I-I). Current models for risk analysis of Complex SoS are insufficient because they often rely on the same risk functions and models used for single systems. By their essence, multifarious characteristics, and emergent nature of the complexity of systems of systems (SoS), cannot be defined, packaged and modeled via the simplicity of our language and models. Rather, we have to develop multiple methods and modeling approaches to better our understanding of their essence, and thus their risk modeling, assessment and management. The key to evaluating the risk to Complex SoS lies in understanding the genesis of characterizing I-I of systems manifested through shared states and other essential entities within and among the systems that constitute Complex SoS. “Essential entities” include the following sample: decisions, resources, functions, policies, decisionmakers, stakeholders, organizational setups, location, and others. This undertaking can be accomplished by building on state-space theory, which is fundamental to systems engineering and process control. We identify two innovations: (i) Building on shared states and other essential entities enables analytical modeling, and thus identifying critical I-I within and among the Complex SoS; and (ii) Reinventing the use of fault trees: The required reliability of each subsystem in the fault tree is replaced with the shared states and other essential entities, representing the I-I that characterize the Complex SoS. Analyzing connections in parallel or in series results in the identification of minimal cut sets. Performing five major case studies has solidified our theoretical and methodological premises on the centrality of shared states and other essential entities in modeling the complexity of SoS, and the process through which to interpret the modeling results.

W2-H.3  11:10 am  What is an effect? Cox LA*; Cox Associates and University of Colorado

Abstract: Much of risk analysis is concerned with calculating how exposures to hazardous substances or activities change the frequency or severity of undesired effects. But what exactly is meant by an “effect” of an exposure? Counterfactual approaches that seek to define effects as differences in frequencies or severities of adverse outcomes with and without exposure, or between higher and lower levels of exposure, are limited by the fact that the difference depends on seldom-specified details of the counterfactual world, such as why and how a hypothetical reduction in exposure occurs. Regression models and other statistical methods for quantifying effects are limited by the fact that no regression coefficient for exposure can represent both the direct effect of exposure (holding all other variables fixed) and the total effect of an exogenous change in exposure allowing other variables (e.g., wealth or income or residential location) to adjust, as well as the exposure-mediated effect of changes brought about by exogenous changes in other variables. A typical regression coefficient represents an unknown mixture of different types of effects (and possibly interactions) and may be useless for predicting how future exogenous changes in exposure alone would affect outcome frequencies and severities. This talk first discusses the impossibility of uniquely defining and measuring effects via regression models or counterfactual models and then proposes an alternative based on predicting the changes in all variables in response to exogenous changes in exposure in causal graph models.

W2-H.4  11:30 am  Concepts and connections, choices and conundrums: the boundary between what is inside and what is outside a risk assessment. Goble R*; Clark University

Abstract: In practice, when an analyst conducts a risk assessment she or he rarely starts from scratch. The purpose of the assessment will constrain what is included and what is not. The purpose, together with past experience is likely to determine the underlying conceptual framing and models. There may appear to be little reason, incentive, or opportunity to question these. Furthermore, almost certainly there will be constraints of time and resources on how much can be included as the analysts struggle to meet their deadlines. There are, however, several practical reasons why analysts should take the time and make the effort to consider and question what possibilities are included and what aren’t. The first and obvious reason is that such questioning may uncover risk pathways that might otherwise be neglected. A second, and only slightly less obvious, reason is that such considerations may also reveal other risks that deserve attention. Even when, after consideration, there appears to be no reason for making significant alterations in the risk pathways and the risks analyzed, the consideration may be used to strengthen the basis for the assessment. After presenting a few illustrations from nuclear safety analysis, other infrastructure risk, emergency planning, and terrorist threats, I will argue that the choice of boundaries is a foundational as well as a practical concern. Boundaries affect both the nature and the strength of the claims that a risk analysis can support. More complicated boundary considerations arise in assessing risks within interconnected systems; the analysis of management possibilities may expose conundrums. Over time “living risk assessments”, risk assessments that are kept up to date as new information and knowledge emerge or new applications are found, should prove valuable. Effective updating will require revisiting the boundaries of the assessments.

[back to schedule]