The influence of visual working memory on correspondence in the Ternus display
16.504, Friday, May 10, 5:30 - 8:00 pm, Vista Ballroom
Elisabeth Hein1, Andrew Hollingworth2, Cathleen M Moore2; 1Department of Psychology, Universty of Tübingen, 2Department of Psychology, University of Iowa
Introduction: Surface feature continuity plays an important role in the perception of an object as a single, persisting entity across change and disruption. For example, color can be used to establish object correspondence across motion, occlusion, and eye movements. However, the underlying mechanisms of feature-based object correspondence are not well understood. Here, we tested the hypothesis that visual working memory (VWM) maintains feature information used to solve the correspondence problem under conditions of ambiguous apparent motion. Method: We used the Ternus display, in which three adjacent discs were presented. After a variable inter-stimulus interval (ISI), the discs were shifted horizontally by one position. Across displays, the correspondence among discs was ambiguous. The discs could be perceived as moving together as a group (group motion) or as one disc jumping across two stationary discs (element motion). Participants indicated the form of motion they perceived. To examine whether the content of VWM modulates the correspondence operation, the discs were assigned colors that could bias the percept either toward group motion or element motion. In addition, participants maintained a color in VWM to prepare for a memory test. The memory color matched either the color of the Ternus discs biasing group motion, the color of the discs biasing element motion, or neither. Results and conclusion: Consistent with previous studies, the proportion of group motion responses increased with increasing ISI. More importantly, the memory color influenced perceived motion. Relative to the no-match baseline, observers were more likely to perceive element motion when the memory color matched the element bias and more likely to perceive group motion when the memory color matched the group bias. These results suggest that feature-based object correspondence is mediated by the content of visual working memory.