Abstract
Studies of spatial representations have typically limited their analysis to memory for object location. Three experiments examined whether another spatial feature, object orientation, could be monitored and represented in a similar fashion. In Experiment 1, an adaptation of the change detection paradigm of CitationSimons and Wang (1998), we found that, whereas unitary location or identity changes were readily noticed, generalized orientation changes were not. Experiment 2 showed that orientation monitoring is strongly affected by layout complexity, viewpoint changes, and the extent of array modifications. Finally, Experiment 3 suggested that an object's behavioral relevance may selectively enhance its orientation processing.
Notes
1In a simple orientation discrimination task (left-right tilt judgments), 6 participants viewed pictures of 12 objects of each category presented in various orientations (± 22.5° from true horizontal or vertical orientations). Using a 1-up/3-down adaptive staircase (CitationGarcía-Pérez, 1998) for each orientation and object combination, task difficulty was manipulated by increasing or decreasing the amount of noise overlaid to the image. These colored Gaussian noise masks obscured a sufficient portion of the objects to render identification impossible, while allowing orientation discrimination to be based on overall shape. Three sets (tools, living, and hybrid) of 8 objects were selected on the basis of perceptual salience, as determined by the amount of unmasked signal necessary to achieve criterion performance. For further details, see CitationMello (2009).