Are the grouping principles of proximity and common region cognitively penetrable?

Purpose: In the past we have found that subjects can detect a pair of identical shapes more quickly when they are located with the same perceptual group than when they are located between different perceptual groups. The purpose of this experiment was to determine whether this effect can be affected by the subjects' strategy for the grouping factors of proximity and common region. Methods: The subjects' task was to respond as quickly as possible whether a row of otherwise alternating circles and squares contained a pair of adjacent circles or squares. The target pair could be located within the same perceptual group or between two different perceptual groups. The critical manipulation was the percentage of within versus between-group trials in a block (25-75%, 50-50%, 75-25%). Results: For groups defined by common region, the difference in reaction time to pairs located within the same perceptual group versus between perceptual groups decreased as the proportion of between-group trials increased. The probability manipulation had very little effect on groups defined by proximity. Conclusions: Grouping induced by common region can be attenuated (although not reversed) by strategy, whereas grouping by proximity appears to be more resistant to strategy. This difference may reflect the intrinsic versus extrinsic nature of the grouping principles: proximity (or location) is intrinsic to the elements in question whereas common region is extrinsic to the elements in question, and thus may be filtered out by attention.