Towards an evidence-based approach to communicating uncertainty in intelligence analysis

Intelligence products have a degree of uncertainty associated with them. This is typically expressed using linguistic probabilities (e.g., 'likely'), and some organizations have adopted standardized lexicons for communicating uncertainty. This paper empirically shows that intelligence analysts use a wide heterogeneity of language to communicate uncertainty. This does not include all of the phrases in standardized lexicons used by the intelligence community. In addition, analysts may use some phrases differently to that advocated. Miscommunication of uncertainty can have deleterious effects on decision-making, and so standardization of uncertainty communication should be evidence-based. This paper discusses ways in which such evidence can be generated.