An Analysis of the Applicability and Trustworthiness of Risk Characterization Approaches

Abstract

The risk field has developed several divergent approaches to risk characterization. These approaches often involve probabilistic reasoning but also incorporate aspects of uncertainties, knowledge, and knowledge strength. This paper develops, reports, and discusses an empirical investigation of some selected common approaches. Based on some derived risk characterization statements, we seek risk analyst opinions on which statements are most preferred, informative, trustworthy, and justified. The results from this investigation clearly show that the most appropriate characterizations include aspects of uncertainty and knowledge-strength judgments beyond pure probabilistic formulations. Overall, the knowledge strength approach has the best scores. While risk characterization components differ across risk regimes, we argue that aspects of knowledge and uncertainty can be prevalent at an explicit and implicit level; and further understanding of those components is essential for further systematization of risk science. We leverage the findings to develop key insights and considerations for future risk research and practice. This paper and insights will be of interest to risk analysts, decision-makers, and policymakers.

Keywords: risk analysis, risk study quality, Data analysis