Abstract
In a stroboscopic induced motion display, a large frame that is instantaneously displaced in a horizontal direction will induce an opposing illusory horizontal motion component in a vertically moving target. Bridgeman & Klassen (1983) demonstrated that this illusory motion is best explained by the induced Roelofs effects, with the displaced frames causing a misperception of target locations. Recently, our laboratory has demonstrated that the induced Roelofs effect is itself caused by a distortion of the subject's apparent midline: a frame, offset from the subject's true midline, serves to pull the subject's apparent midline in the same direction (Dassonville & Bala, VSS 2002). Taken together, these findings suggest that a stroboscopic induced motion display can be used to conveniently measure the time course of the apparent midline distortions that underlie the Roelofs effect. Subjects viewed a large frame that was repeatedly displaced left and right, in a square-wave fashion, with a frame duration of 600 ms at each position. Within this moving frame, a short duration (16.6 ms) target was displaced vertically at the same frequency. In separate trials, target onsets were timed to occur after a given delay (0 – 550 ms) from each displacement of the frame. Subjects were required to adjust the horizontal positions of the upper and lower target positions until the induced motion was cancelled (i.e., the targets appeared to be moving in a purely vertical direction). Thus, the true displacement of the targets provided a measure of the magnitude of induced motion, which could be used to subsequently compute the magnitude of the frame-induced distortion of the apparent midline. We found that each displacement of the frame caused a slow shift in the apparent midline, moving from one extreme to the other in approx. 300 ms. Surprisingly, this distortion of the apparent midline occurred in a predictive fashion, with changes beginning approx. 200 ms before the frame displacement.