Abstract
Recent research investigating visual working memory (VWM) crowding showed that crowding is observed for VWM representations similarly to visual perception. When the VWM arrays are presented in an array with a middle target and two flankers at the edges of the array, accuracy is worse than when an edge item is the target (Yörük et al., Psychonomic Bulletin & Review, 2020). However, due to the sequential nature of the VWM crowding task, the most recent item in the sequence showed a strong recency effect, and might have been processed free from crowding (Yörük & Tamber-Rosenau, VSS, 2020). In the present study, we tested whether the introduction of visual masking would mitigate the recency effect for the last presented item. Moreover, we expected to observe particularly strong crowding with masking (Vickery et al., Journal of Vision, 2009). Three oriented bars in a radial configuration relative to central fixation were presented sequentially. A visual mask followed this array, and the participants were asked to report the orientation of one of the items in the sequence. Results showed that there was a significant interaction between the spatial and temporal positions of the target item for the accuracy of orientation reports, and there was a significant recency effect despite the visual mask. According to post-hoc analyses, when the target item was presented in the final temporal position, there was no difference between the accuracy of spatial middle and edge targets. However, crowding was the strongest (greatest error) when the target was in the middle ordinal position, with significantly higher errors for the middle target than the edge targets. Thus, while masking did not eliminate the recency effect, it yielded an interaction between spatial and temporal VWM crowding.