Alma College Psychology Department
The Effects of Different Orientations on the Strength of the Zollner Illusion
Erin Goodrich 00ECGOOD@Alma.edu
Jennifer Dunham 99DUNHAM@Alma.edu
The Zollner Illusion is traditionally a series of parallel lines intersected by transversals, at an angle of 30 or 40 degrees. These transversals make the parallel lines appear to converge towards and diverge away from eachother (see Figure 1). By alternating the direction that the transversals are slanting in, it makes the parallel lines appear to slant in alternating directions, as well. The Zollner Illusion is traditionally tested in either a horizontal or vertical position. Whether on the grounds of learning to adjust to the forces of gravity or because of the perhaps significant arrangement of neural stucture, it has been proposed that these two positions are particularly stable. (Maffei and Campbell, 1970) Horizontal and vertical lines would, therefore, resist distortion more than lines in other positions. The subsequent experiment was designed to test whether the strength of the Zollner Illusion is effected by its orientation.
Figure 1: Zollner Illusion
Methods
Twenty male and female college-aged students were recruited for the experiment. The experiment was designed using the Eye Lines software developed at Alma College. (Beagley, 1990) The computer screen was tilted at a 30 degree angle to ensure that the stair step effect, created by the pixels at any angle other than gravitational vertical and horizontal, did not effect the Ss ability to correctly adjust the lines. There was a 5 x 6 7/8 inch vertical square cut out of a piece of posterboard and placed over the screen for the first trial and one of the same size cut out at a 30 degree angle and placed over the screen for the other trial, to act as a frame for the screen. The Zollner Illusion used in the experiment consisted of two parallel lines, one of which was intersected by 7 transversals each at a 30 degree angle from the line. The other parallel line was adjustable. The experiment was set up using three different stimuli. The first stimulus was positioned at a 30 degree angle to compensate for the screen being tilted and therefore was horizontal. The second one was set up at a 120 degree angle, appearing vertical. The third one was set up at a 75 degree angle in order to look like it was slanted at a 45 degree angle. Each stimulus was presented twice for each of the two trials. The Ss were asked to move the adjustable line, using the mouse on the computer, until it appeared to be parallel to the nonadjustable line. Each subject ran through the experiment twice, once with the vertical posterboard background and once with the slanted posterboard background.
Results
As shown in Figure 2, the parallel lines placed in the vertical position produced the weakest illusion, followed by those in the horizontal position, with the 45 degree orientation showing the strongest illusionary effects. There was no real difference between the results collected with the vertical background and those with the slanted background, showing that the surrounding environment did not impact the Ss ability to perceive each of the three stimuli.
Discussion
The Zollner Illusion can be explained by the theory of the regression to right angles. This theory contends that there is a tendency to overestimate accute angles and underestimate obtuse ones (Hotopf and Ollerearnshaw, 1972), making the line and its transversals appear more perpendicular to eachother than they really are. This in turn causes the transversed line to appear slanted. By the way the transversals were positioned, we hypothesized that the Ss would overadjust the line in the vertical and 45 degree orientations and underadjust the line in the horizontal orientation. The data collected supported this hypothesis. It has also been proposed that gravitational horizontal and vertical are particularly stable positions (Maffei and Campbell, 1970). We then hypothesized that the Zollner Illusion would have the least effect on the Ss when in the horizontal or vertical position and more of an effect when slanted at a 45 degree angle. As shown in Figure 2, the average results for the vertical Zollner Illusion (Trial 3) are the closest to parallel, followed by the average results for the horizontal Zollner Illusion (Trial 2), with the average results for the 45 degree Zollner Illusion (Trial 1) being the furthest from parallel, thus supporting our hypothesis and suggesting that horizontal and vertical lines resist distortion more than lines in other positions. More experimentation would have to take place to discover whether this stability is due to out learning to adjust to the forces of gravity, the arrangement of neural structure, or some other as of yet unforeseen determinant.
References
Beagley, W. K. (1990) Eye Lines [Computer Program]. Alma, MI: Alma College.
Hotopf, W. H. N. and Ollerearnshaw, C. (1972). The Regression to Right Angles Tendency and the Poggendorff Illusion. Br. J. Psychol. 63, 3, 359-367.
Maffei, L. and Campbell, F. W. (1970). Neurophysiological Localization of the Vertical and Horizontal Visual Coordinates in Man. Science, N.Y. 167, 386-387.
More Posters
Web Posters from other courses and projects
Alma College Psychology Department