Robert Cormack, Randolph Blake, and Eric Hiris (1991) discovered an illusion that can change the perceived motion of a figure seen peripherally up to 90 degrees. They determined that the illusion is not caused by the so-called aperture problem or by a Hermann grid component. Their most probable explanation involves a series of vectors which add up to create the illusion. However, they propose five vectors which do not add up to more than a small number of degrees of difference from the actual motion. A simpler version of the same illusion that contains only two or three vectors may do a much better job of explaining how the illusion can be stronger than just a small number of degrees.
MethodsSubjects Subjects for this experiment consisted of 12 Alma College students.
Apparatus Subjects viewed a HyperCard animation on a Macintosh computer. The animation moved at 2 pixels per frame and 15 frames per second. The animation consisted of a square (44 pixels) moving at 45 degrees up and to the right from the bottom left corner of the screen (see figure 1). The box travels to approximately the middle of the screen for five cycles. The box contained three black lines that moved the same speed upward as the box moved diagonally (2 pixels per frame). Another computer using Eye Lines was set up so that the subjects could adjust a line to any angle.
Procedure Subjects viewed the stimuli from approximately 2 feet from the computer screen. They were asked to stare directly at a thumb tack that was placed on the wall approximately 2 feet above and behind the bottom of the computer screen. Once the animation was complete, each subject was asked to adjust a line using a mouse to the angle that the box appeared to be moving. After three of these trials, the subjects were asked to stare directly at the box and repeat the process of adjusting the line for three more trials.
When subjects viewed the animation peripherally, their mean error was about .9 degrees. When subjects viewed the animation foveally, their mean error was about -2.2 degrees (see Figure 2). The difference between peripheral and foveal is 3.1 degrees.
These results are disappointing. It appears that the simple vectors model does not create an illusion at all. However, this is not conclusive evidence against it. Future research will try placing the monitors close enough together for the subjects to be able to make a more rapid enough judgement of the angle of the box's direction. Also, the influence of corrective lenses will be taken into account. Thirdly, the adjusting line will be contained inside of a 90 degree angle to give subjects a gauge by which to determine the angle of the box. Finally, subjects will be seated in an adjustable chair to ensure that they view the monitor from the same height.
ReferencesBeagley, W. Eye Lines. [Computer Program]. Alma, MI: Alma College.
Cormack, R., Blake, R., Hiris, E. (1992). Misdirected visual motion in the peripheral visual field. Vision Research, 32, 73-80.