Abstract
In multiple object tracking studies, people can track the locations of 4–7 moving objects. Additionally, it has been hypothesized in the object-file literature that as an object moves, the association between object position and object surface features is automatically updated. However, people are often poor at recalling non-spatial features associated with tracked objects (Pylyshyn, 2004). In the present study, we sought to determine the capacity for updating the association between object location and surface features. In Experiment 1, participants saw four boxes and tracked either one, two, three, or four. Colors appeared in each box, the colors were removed, and the boxes moved to new locations. One of the tracked objects was cued, and the participant reported which of the four colors had been associated with that object before motion. Participants' updating capacity was approximately two color-position bindings. In Experiment 2, we replicated these results using a change detection task. Again we found that participants could update color-position binding for approximately two objects. This updating capacity was significantly lower than capacity for color in a standard VSTM task that did not require spatial updating. In Experiment 3, the stimuli to-be-updated contained two features. Colored shapes appeared in the boxes, and after motion, participants were asked to report color alone, shape alone, or either shape or color. Capacity for updating two features of an object was no lower than capacity for updating either feature alone, suggesting that spatial updating is constrained by the number of objects rather than the number of features, as found in standard VSTM tasks (Luck & Vogel, 1997). Together, these results indicate that VSTM is used to update the binding of surface features to locations in an object-based manner. However, this updating can be accomplished only for a subset of objects in VSTM.