In 1991, Robert Cormack, Randolph Blake and Eric Hiris found an illusion in which the direction of motion of an object viewed peripherally moving against a background textured by parallel lines may be misperceived by as much as 90 degrees. The movement had to be viewed peripherally, or the illusion would disappear. In the original illusion, a textured background of solid black bars set from left to right at a 45 degree angle are passed over by a horizontal black bar moving from bottom left to top right. When viewed peripherally, the bar appeared to be moving straight up. In their report, Cormack et al. stated that "the distance travelled is not critical." However, several observers noticed that the illusion faltered when the moving bar reached the edge of the textured pattern. So, how long could the illusion really be sustained if the textured background covered a greater area? Would the illusion still fail after a certain distance travelled?
Our subjects were twelve Alma College undergraduates. Our stimulus was similar to that of Cormack et al., with an area of solid black parallel bars set from left to right at a 45 degree angle passed over by a horizontal black bar moving from bottom left to top right. Our area of textured background spanned the entire seventeen-inch computer screen. Starting at the bottom left side of the screen, we made our bar move five varying distances horizontally on the screen: two inches, four inches, six inches, eight inches and ten inches. Each test was run three times, so we ended up with thirty-six sets of data for each of the five tests. Using the program, Eye Lines for Macintosh, each participant was then asked to adjust the angle of a line on an adjoining monitor to show which angle they perceived the bar to be moving. The program records the error relative to the true direction of movement.
Mean illusion magnitude was the same for all distances. As shown in Figure 1, the amount of time the illusion is viewed does not make a consistent amount of difference in the strength of the illusion. However, as the experiment proceded, subjects not only described the bar moving upward but also reported the bar moving in circles and even backward. It appeared that longer trials caused more cirular and backward movements.
Figure 1. This chart shows the minimal change in the illusion between the five distances travelled.
There was not much research to be found on this illusion since the original research in 1991. So there was freedom to stretch the illusion further and see what we found. It seemed logical that at some point in motion the true orientation of the moving bar would be detected, since change in location corresponds as movement at some point. While we didn't find any significant results to support our hypothesis, we did gather some very strange reactions from participants. As the experiment proceded, subjects not only described the bar moving upward but also reported the bar moving in circles and even backward. It appeared that longer trials caused more cirular and backward movements. In this way the illusion was not broken but altered, since circular movements do not correspond with the illusion of moving in a straight line. Since we were asking our participants to judge the path against an adjustable straight line, we do not have any feasible data on any motions that deviated from that path. Also, many participants claimed that having an area that large covered with a background of diagonal black and white lines made the illusion difficult to look at.
Cormack, R., Blake, R., Hiris, E. (1991). Misdirected Visual Motion in the Peripheral Visual Field. Vision Research, 32, 73-80.
Beagley, W. K. (1990) Eye Lines [computer program]. Alma, MI: Alma Colege.
More Posters
Web Posters from other courses and projects
Alma College Psychology Department