Poster B134, Sunday, March 26, 8:00 – 10:00 am, Pacific Concourse
Deconstructing a Superadditivite Effect of the Sander Parallelogram Illusion
Robert L. Whitwell1, Sean Olsen1, James T. Enns1; 1The University of British Columbia
One contested issue in the debate on whether or not action resists pictorial illusions is centred on how attention operates in perception and action tasks. The posterchild for this issue is the well-known Ebbinghaus illusion, which is most famously viewed with the two sets of surrounding inducers and their respective target inner circles (the ‘dual’ configuration). Critics have claimed that the perception-action dissociation depends on spatial attention being deployed to two locations simultaneously or in rapid succession: the dual configuration yields an effect that is greater than the sum of its components (superaddivity) when the sizes of the inner disks are adjusted simultaneously but not when the adjustable stimulus (e.g., the participant's grasping hand or a third disk) is isolated. Here, we test the Sander Parallelogram illusion for superadditivity. In experiment 1, we quantified the effect of the Sander illusion across three different methods of adjustment in which the comparator is isolated: line-length matching, gap-length matching, and line-length to thumb and finger aperture size (manual) matching. We found a reliable and powerful effect of the illusion (16.5%) across all three methods. In experiment 2, we tested for superadditivity by contrasting the composite Sander display against the sum of its two component parts. We found a clear superadditive (5%) boost for the composite with no differences across measures. Thus, despite the isolation of the adjustable comparator which rendered simultaneous comparison moot, the composite Sander Parallelogram illusion induces superadditivity. Taken together, our results highlight the scene-based nature of spatial attention and vision-for-perception.
Topic Area: PERCEPTION & ACTION: Vision