Abstract
Most studies of pictorial illusions rely on session-wise averages – an approach that assumes that effects of interest remain constant across successive iterations of the response within a test session. Recent evidence, however, suggests that this assumption is not always justified. For example, our group has shown that the illusory influence of the Ponzo display on the apparent size of objects is mitigated with practice for ‘awkward’ grasps executed with the ring finger and the thumb of the right (dominant) hand over the course of three consecutive days of testing, but not for awkward grasps executed with the left (non-dominant) hand (Gonzalez et al., 2008). Notably, substantial within-session reductions in the illusory effects were observed for grasps executed with the right hand. Could mechanisms underlying the within-session reduction in illusory effect extend to ‘precision’ grasps executed with the thumb and forefinger? To answer this question, we asked participants to manually estimate the length of single targets with their thumb and finger aperture a matching amount or to grasp them in a blocked ABA design. In both tasks, participants’ responses were correlated, with equal sensitivity, to target length, but only manual estimates remained consistently biased by the illusion. In contrast, the illusory effect on grasps decreased linearly over the course of testing. The consistency of the effect of the illusion across manual estimates cannot easily be attributed to a lack of haptic feedback, since the manual estimates of an additional group of participants were consistently affected despite picking up the same target immediately following each estimate. We offer three possible explanations that can account for our findings: low-level motor calibration, motor learning, and another that invokes cognitive or attentional set-shifting – a process through which the saliency of stimulus and task features is updated following a switch in task demands.
Meeting abstract presented at VSS 2012