Marianna Barcenas
The London School of Economics
This paper explores the success of Brian Hill’s (2019) non-Bayesian framework for practical decision-making under ambiguity, where ambiguity is taken to refer to situations where an agent is offered partial information from which they may make inferences, but are unable to build a complete picture of the involved risks. Ambiguity and ambiguity-averse behaviour, are present in many critical decision problems, including medicine and climate policy; therefore, it is vital that ambiguity is included in a normative framework for decision. Hill’s framework uses confidence in assigned degrees of belief to allow for comparisons between ambiguous or imprecise probabilities. Essentially we should choose to act considering the degree of belief for which we have the most confidence in, with confidence being formally quantifiable. I argue that the success of Hill’s framework depends on how clearly confidence can be communicated between all parties involved in the decision-making process, which is crucial if there is to be alignment between judgements, values and choices in policy decisions. I argue further that the lack of an accepted theory for scientific confirmation is a potential cause for breakdowns in communications within policy decision making and briefly offer some potential practical frameworks for scientific confirmation that can work alongside the confidence model.
Chair: Niklas Parwez
Time: September 11th, 16:20 – 16:50
Location: SR 1.004
SOPhiA 2026 