Abstract
Visual Working Memory (VWM) is an online workspace that holds a limited amount of information, about 3-4 objects, in an active state for a short period of time (Luck & Vogel, 1997). A way to cope with such capacity limitations is to group and parse information into integrated units in VWM. Gestalt grouping cues, such as illusory objects, have previously been shown to improve VWM performance (Allon et al., 2018, Gao, et al, 2016; Peterson, et al, 2015). While illusory objects clearly lead to behavioral improvements, the question of how these improvements arise remains unanswered. Here, we ask whether the behavioral benefits derived from an illusory object (eg., a Kanizsa triangle) result from a reduction in storage demands in VWM. We recorded EEG while subjects performed a bilateral change-detection task, with memory arrays configured in four conditions: 1 item, 3 items forming a Kanizsa triangle, 3 proximity grouped items, and 3 ungrouped items. The number of items held in VWM was assessed via the contralateral delay activity (CDA). We found a behavioral performance benefit for the Kanizsa triangle condition, replicating other studies. However, the CDA amplitude for the Kanizsa condition was not significantly different from the 3 proximity grouped and 3 ungrouped item conditions. These results suggest that the behavioral benefits derived from the illusory object Kanizsa triangle are not a result of reduced storage demands in VWM. Future follow-up studies will aim to further delineate how the behavioral benefits of the Kanizsa triangle arise.