Measuring Object-Centred Attention Based on a Cueing Paradigm

Visual attention can be allocated to a location in visual space and/or to a representation of an object in the visual field, independently of their spatial location. In Posner's cueing paradigm, it is assumed that attention is moved to and then engaged at a cued location. If the target appears at an uncued location, attention first has to be disengaged from the cued location before moving to and engaging the target location.

On the basis of this paradigm, we designed an experiment to measure the disengaging of attention from objects, independently of location. For this purpose we used the bistable Necker cube that can be perceived as two different object configurations, depending on the position of the front side of the cube (lower left or upper right). The subject was instructed to react when he perceived the Necker cube as a previously presented model configuration, ie a stable cube. Each condition started with a bistable cube (with equal luminance of the ribs) that gradually evolved into the model configuration, by manipulating the luminance of the ribs of the cube. Prior to this the subject was cued by a cube similar to the model in the valid condition and cued by a dissimilar cube in the invalid condition.

The results showed a significant difference (F_1,10=7.35; p<0.05) between the valid and invalid cue condition, indicating a significant cost for the invalid cue condition. This effect is in accord with the well-known set effect found previously with bistable figures. The difference between the valid and the invalid cue conditions will be interpreted as disengaging attention from an object. Thus, object-based components of attention can be examined by paradigms similar to Posner's paradigm for location-based attention.