New Work on Negative Ranks and How We Handle Uncertainty

Hanbin Go has been working for the last couple years looking at alternative accounts of how we deal with uncertainty. The traditional account is that we use our subjective probabilities. However, the fact that we are so bad at providing objectively correct probabilities, or estimating randomness, and tricked by so many fallacies suggests that this probabilist account is at the best incomplete.

There are extremely formal (in the mathematical sense) and normatively robust alternatives. Wolfgang Spohn has spend a 50 year career looking at negative ranks as the right way to deal with belief and its updating. Hanbin and I could not find very many attempts to apply this philosopher's theory to human decision making, and so Hanbin has spent the last couple of years developing some methods. Can people use ranks to express their belief or surprise in propositions? Are these expressions consistent and robust across propositions and measurement methods, e.g. online or inperson; real-life situations or observed event frequencies; and how do the ranks provided relate, if at all, to the objective probabilities or subjective reports? A lot of questions and a lot of data. One of Hanbin's first comprehensive presentations of this data will be this month at the CSBBCS held in Guelph. If you missed the presentation you can find a link to his poster here.