False predictions about the detectability of visual changes: the role of beliefs about attention,memory, and the continuity of attended objects in causing change blindness blindness

Recently, a number of experiments have emphasized the degree to which subjects fail to detect large changes in visual scenes. This finding, referred to as "change blindness," is often considered surprising because many people have the intuition that such changes should be easy to detect. documented this intuition by showing that the majority of subjects believe they would notice changes that are actually very rarely detected. Thus subjects exhibit a metacognitive error we refer to as "change blindness blindness." Here, we test whether CBB is caused by a misestimation of the perceptual experience associated with visual changes and show that it persists even when the pre- and postchange views are separated by long delays. In addition, subjects overestimate their change detection ability both when the relevant changes are illustrated by still pictures, and when they are illustrated using videos showing the changes occurring in real time. We conclude that CBB is a robust phenomenon that cannot be accounted for by failure to understand the specific perceptual experience associated with a change.     

Change Blindness Blindness: The Metacognitive Error of Overestimating Change-detection Ability

Recent research has demonstrated that subjects fail to detect large between-view changes to natural and artificial scenes. Yet, most people (including psychologists) believe that they would detect the changes. We report two experiments documenting this metacognitive error. In Experiment 1, students in a large General Psychology class were asked if they thought they would notice the change in four different situations previously tested by Levin and Simons (1997) and Simons and Levin (1998). Most claimed that they would have noticed even relatively small changes that real observers rarely detected. In Experiment 2, subjects were tested individually and half were asked to predict whether someone else would detect the changes. Subjects again overestimated the degree to which changes would be detected, both by themselves and by others. We discuss possible reasons for these metacognitive errors including distorted beliefs about visual experience, change, and stability.