Hindsight bias in judgments of the predictability of flash floods: An experimental study for testimony at a court trial and legal decision making

As part of the first author's expert testimony at a court trial, we investigated hindsight bias in perceptions of the predictability of a real flash flood. Participants were presented with pictures taken before the flash flood and asked to rate the muddiness of the water and judge the likelihood of flooding in Experiment 1. Participants who were informed that a flash flood had occurred perceived the river as muddier and judged a flood to be more likely than control participants. Their judgment was biased by hindsight. The results of Experiment 2 revealed that when the causality from muddiness to flooding was instructed, participants judged the river as muddier. These results were interpreted as top‐down perception being implicitly adjusted by outcome information and causality information. The judges decided that the defendants had no responsibility to predict the flash flood, taking the hindsight bias into consideration.