Confidence in risk assessments

Jonathan Rougier*

*Corresponding author for this work

Research output: Contribution to journalArticle (Academic Journal)peer-review

1 Citation (Scopus)
59 Downloads (Pure)

Abstract

Risk is assessed with varying degrees of confidence, and the degree of confidence is relevant to the risk manager. This paper proposes an operational framework for representing confidence, based on an expert’s current beliefs about how her beliefs might be different in the future. Two modelling simplifications, ‘no unknown unknowns’ (NUU) and an homogeneous Poisson process (HPP) make this framework trivial to apply. This is illustrated for assessing an exceedance probability for a large event, with volcanic risk as a specific example. The paper ends with a discussion about risk and confidence assessments for national-scale risk assessment, including several further illustrations.
Original languageEnglish
Pages (from-to)1081-1095
Number of pages15
JournalJournal of the Royal Statistical Society: Series A
Volume182
Issue number3
Early online date19 Apr 2019
DOIs
Publication statusPublished - 1 Jun 2019

Keywords

  • Exceedance probability
  • Prospective interval
  • Reasonable worst case
  • ‘Bazaar of experts’
  • ‘No unknown unknowns’

Fingerprint Dive into the research topics of 'Confidence in risk assessments'. Together they form a unique fingerprint.

Cite this