Abstract
Object substitution masking occurs when attention is distributed over a large region, and a sparse, non-overlapping, and temporally trailing mask impairs the perception of a briefly presented object (Enns & Di Lollo, 1997). Previous work has shown that masking is stronger if mask and target are grouped by color or motion. This led to a hypothesis that when the mask and target are within the same “object file”, the trailing mask “overwrites” the file so only the mask is easily perceived (Lleras & Moore, 2003; Moore & Lleras, 2005). However, it is unclear whether this holds for all types of grouping. We investigated whether good continuation and containment grouping between target and mask affects masking strength. We presented subjects (n = 12) with a ring of 8 items for 30 ms, where the target was cued by a four-dot mask, with simultaneous or 320 ms delayed offset, in two experiments. In the collinearity experiment, the mask was either square or rotated to form a diamond. Subjects reported the orientation of the target: a pair of horizontal or vertical lines which were collinear with the square mask, but not the diamond mask. In the containment experiment, the four-dot mask was either inside or outside the target: a spiky or wavy circle. Collinearity relieved masking, with a masking effect (% correct simultaneous - % correct delayed offset) of 8% (91%–83%) for the collinear mask, and 17% (91%–74%) for the non-collinear mask (p = .0002). Additionally, masking was stronger (p = .0419) when the mask was inside, 18% (85%–67%), compared to outside, 12% (72%–60%). Grouping was stronger when target and mask were collinear, and when the mask was inside a circular target. However, these strong grouping conditions respectively produced weaker and stronger masking. This suggests that the type and strength of grouping between target and mask has a complex relationship with mask effectiveness in object substitution masking.