Selective storage in visual working memory: Distinguishing between perceptual-level and working memory-level mechanisms

The visual working memory system is dramatically limited in its small storage capacity (34 items) and its relatively slow process of encoding items into memory. This low-capacity system is in stark contrast to the high-capacity perceptual system that is capable of identifying a large number of objects very rapidly. Despite the processing differences of these two systems, they must directly interact to store perceived information in memory. This capacity difference necessitates some form of mechanism to select only a small number of perceived objects for storage into visual WM. This mechanism would protect the limited resources of the visual WM system from being overloaded with nonessential objects, by allowing only the most relevant items to be stored in memory. In this dissertation, I will provide evidence for the existence of such a mechanism. A sequence of experiments will be described that will demonstrate and characterize this attention mechanism by examining the ability of observers to selectively retain items in visual WM. In addition, this visual WM-level attention mechanism will be differentiated from a traditionally studied form of attention that is thought to operate at an earlier stage in processing.