Knowledge Bank

Working memory (WM) is an important cognitive function that maintains only a finite amount of information that is no longer externally available. Because of WM's capacity limitations, information storage optimization is essential. One optimization method occurs through grouping, where items appearing to be parts of a whole are "chunked" for more efficient storage. An example is the illusory object effect of visual WM, where the orientation and alignment of separate items' contours create the appearance of a composite object, shown to allow more efficient storage. However, it is unclear whether the benefits of storing illusory objects as composites are automatically derived or dependent on individual strategic use. In our study, we manipulate the degree to which participants can expect the illusory object effect to be useful and measure the degree to which objects are stored compositely. Participants completed a task where objects were briefly shown on screen prior to a test of one object's color or orientation, with the frequency of each test type manipulated between trial blocks. Subjects were randomly divided into two groups where one group received stimuli only in illusory object-forming configurations and the other received stimuli only in grouped but randomly oriented configurations. Within the illusory object group, performance was significantly higher on orientation tests regardless of within-block frequency. Color and orientation recall were similarly high when each test was likely, but color test accuracy dropped when tests were unlikely. Within the random orientation group, subjects performed significantly better on tests of color than of orientation. Test accuracy for orientation dropped with test likelihood, but only a slight drop was observed for color. Between groups, performance was significantly different on orientation but not color tests, regardless of test likelihood. Our results are consistent with past studies, exhibiting an illusory object memory benefit limited to related object features. However, our findings suggest a lack of modulation by expectation or strategic use, and that the illusory object effect may occur automatically. Our observations provide insight into the storage and optimization processes of WM, though further testing is needed to elucidate precisely when and how grouping processing takes place.