The results were clear. If an event had actually occurred, people exaggerated the probability that they had assigned to it earlier. If the possible event had not come to pass, the participants erroneously recalled that they had always considered it unlikely.
