Abstract
In a series of experiments, we investigated how retrieving object features affects the precision of memory for their location. Participants viewed 6 objects randomly located on a display for 2.5 seconds. Each object was made up of a combination of one of 6 colors, 6 shapes, and 6 patterns for a total of 216 possible combinations. After a delay of approximately 4 seconds, participants were shown one feature (color, shape, or pattern) and asked to indicate the location on the screen of the object that contained that feature. After responding, participants were asked to indicate another feature the object contained. Participants were significantly more accurate to indicate the location of the object when they were given a color property, than a shape, or pattern. In addition, location accuracy was positively correlated with feature accuracy. In a second set of experiments, 3 of the objects had the same color, shape, and pattern (duplicate objects), and 3 had unique combinations (unique objects). Memory for spatial location and features was significantly more accurate for duplicate than unique objects. The overall spatial accuracy of both the 3 unique and 3 duplicate objects was also significantly higher than 6 unique objects. This indicates that participants may be able to chunk together similar objects to reduce total memory load. These results have important theoretical implications for how feature and spatial information are bound and used to reference each other.
Meeting abstract presented at VSS 2015