For example if they had predicted that their likelihood of being robbed was 30%, but the average likelihood was actually 20%, they might adjust their estimate to 22%.

FORBES: Pollyanna Had Brain Damage! Study Claims Faulty Brain Wiring Sparks Unrealistic Optimism

If the information was worse than expected, then they tended to adjust their estimate much less (if the average likelihood was really 40%, and they estimated 10%, they might adjust to only 15%), which suggests that they were largely ignoring the new information.

FORBES: Pollyanna Had Brain Damage! Study Claims Faulty Brain Wiring Sparks Unrealistic Optimism