The Advantage of Numeric Uncertainty Information For Decision-Making With Two Thresholds

Abstract

Evidence suggests that people make better decisions based on forecasts with numeric uncertainty estimates (e.g., 30%) compared to single value (deterministic) forecasts. Much of this work has focused on binary tasks. However, most real-world situations include more array of options. Advantages for uncertainty information may not be seen with more options because of the increase in cognitive load. The present study aims to answer whether the advantage for probabilistic compared to deterministic forecasts persists in situations with three options. To investigate, a school closure paradigm with three-option (close, delay, open) was compared to two-option condition (close, open). The results suggest that the advantages for probabilistic forecasts hold across task complexity. Moreover, participants in two-option condition showed greater proportion of risk-seeking errors than risk-averse errors, whereas the tendency was reversed in three-option condition. The study suggests that numeric uncertainty information can be beneficial in complex situation.