Reappraising the object effect: the role of attentional shift across a non-uniform region
Degree GrantorUniversity of Canterbury
Degree NameMaster of Arts
Object-based attention refers to the facilitation in visual processing when stimuli belong to the same object compared with when they belong to different objects. In a typical experiment (e.g., Egly, Driver & Rafal, 1994), participants see two objects (e.g., two rectangles), then an informative spatial cue at one end of an object, followed by a target at either the same location or a different location. When the cue and the target appear at different locations, the target appears either at the other end of the same object (the same object condition) or in a different object (the different object condition). Importantly, the spatial separation between the cue and the target is identical in the same and different object conditions. The typical result is that the participants are faster and/or more accurate in the same object condition than in the different object condition. This object effect is taken as evidence for object-based guidance of attention.
In general, object effects are found more reliably in studies that use a spatial cuing paradigm, in which the target is preceded by an informative spatial cue as described above, as compared to studies that use a feature comparison paradigm, in which two stimuli are presented simultaneously and the task is to compare whether they are the same or different. In a recent study, Chen and Cave (in press) conducted a series of experiments using a feature comparison task. They found no object effects when the orientation of the target configuration was controlled. Their study raised the question about object-based allocation of attention in feature-comparison tasks.
The present study examined the cost of shifting attention across a uniform vs a non-uniform region as a function of the orientation of the target configuration, and the degree to which the object effect reported in previous research can be explained by the cost incurred when shifting attention across a non-uniform region. In three experiments, participants saw letter targets presented on one or two objects, and the targets were aligned either horizontally or vertically. The task was to determine whether the targets were the same or different. In Experiment 1, targets were presented simultaneously within an object, and the shape of the object was manipulated. The goal was to examine the processing efficiency of the targets that were separated by a uniform region or a non-uniform region when no shifts of attention were required. Responses were faster and more accurate when the targets were horizontally configured than vertically configured, indicating a horizontal benefit. No difference was found between the uniform and the non-uniform conditions. In Experiment 2, targets were presented sequentially within the same object, and they were either across a uniform or a non-uniform region. A horizontal benefit was again found. In addition, performance was impaired in the non-uniform condition compared with the uniform condition. In Experiment 3, the targets were again presented sequentially. In some trials, two objects were presented in the target display, and the targets were either within the same object with a non-uniform region between them, or between two different objects. Once again, a horizontal benefit was found. Importantly, no difference in performance was found between the same and different object conditions. Taken together, these results suggest that the object effects in previous spatial cuing studies may be caused by shifting or spreading of attention across a non-uniform region rather than object-based guidance of attention. The results also provide converging evidence to the findings of previous research, which show that the horizontal benefit is a robust phenomenon, and that shifting or spreading attention across a non-uniform region incurs an additional cost compared to shifting or spreading attention across a uniform region.