Abstract
The deployment of attention within a complex scene has Received much attention in the last decade (e.g., O'Regan, Rensink, & Clark, 1999; Rensink, 2000; Simons & Levin, 1998). Much of this work demonstrates a startling “blindness” to changes in objects, arrangements, even people. Some work suggests that tacit social goals may guide the deployment of attention (Simons & Levin, 1998). We examined such a possibility. We hypothesized that the movements of people would garner attentional resources since such movements might have social significance to an observer.
To test this proposal, we compared the detection of changes in a person's movement to the detection changes in a moving person. We constructed two short videos. In each video, we created five abrupt, unusual changes, two to the appearance of an actor and three to the behavior of one or more actors. A change of appearance, for example, was accomplished by changing an actor's clothing from dark to light as she passed behind a pillar. A comparable change in behavior involved an actor walking behind a tree and then reappearing with a limp. In every target change, the actor moved throughout the scene.
We presented 50 undergraduates one of two videos and then asked them to recall any surprising changes in the video. Two assistants independently coded observers' responses. Responses were regarded as reflecting detection of the target change if the observer identified either the person involved, the place where it occurred, or the change itself. As expected, the frequency of detection varied with the type of change, F(4, 98) = 4.34, p < .01. Pairwise comparisons suggest that observers were more likely to report changes in actors' behavior than in their appearance. Subsequent probes demonstrated that observers were capable of detecting and remembering all of the target changes. These findings suggest that the movement of entities in a complex scene may be more salient than the appearance of moving entities.
Supported by NEI grant R01 EY12300