Two Faces of Subjective Uncertainty

In my talk I will argue that people carry with them dual intuitions concerning the nature of uncertainty.   In some cases they attribute their uncertainty to deficiencies in their own knowledge, information, or mental model of the event in question (“epistemic uncertainty”); in other cases they attribute their uncertainty to causal systems in the world that exhibit stochastic properties (“aleatory uncertainty”). I will review some support for this thesis from prior literature and then present a series of studies showing that people make such intuitive distinctions.  First, I will show that epistemic uncertainty is marked in natural language by statements such as “I am 80% sure that…” or “I am reasonably confident that…” and aleatory uncertainty is marked by statements such as “I think there is an 80% chance that…” or “I believe there is a high probability that…”.  Second, I will show that people can reliably distinguish epistemic from aleatory uncertainty in their rating of events, and that these ratings predict evaluations of whether people who make correct/incorrect predictions are deserving of credit/blame and/or are seen as lucky/unlucky.  I will also show that hindsight bias is driven partly by shifts in these ratings before versus after uncertainty is resolved, and that people with lower self-esteem make self-serving attributions concerning the nature of uncertainty.  Moreover investing behaviors (time horizon and diversification) are associated with individual differences in ratings of market uncertainty.  Third, I’ll show that people tend to make more extreme likelihood judgments, holding evidence strength constant, for events that are deemed more “epistemic” and less “aleatory,” and this may lead to greater overconfidence.  I’ll conclude with a discussion of further implications and directions for future research.