Effect of Additional Lines on the Jastrow Illusion
According to the assimilation theory, when two objects in a visual array are compared in size, an individual will perceive each object as the average size of the other objects involved (Pick & Pierce, 1993). This theory can be seen in phenomena such as the Jastrow illusion. This illusion has two same-sized crescents either stacked on top of one another, or placed side by side. In each instance, one crescent - either the bottom one or the one on the right, depending on how the objects are stacked - appears larger than the other. The assimilation theory would explain this by saying that since the longer arc of the second crescent is placed next to the shorter arc of the first, the entire second crescent will appear longer (Pick et al. 1993). The Jastrow is an example of a size constancy illusion.
A second illusion that illustrates this size constancy is the Oppel inverted T. In this illusion, two same-sized bars are placed perpendicular to one another, to make the shape of a T. However, the vertical line appears longer. This phenomenon can be explained by a theory proposed by Oppel that a line split directly in the center will look shorter than a line of the same length (Erdfelder & Faul, 1994). The purpose of this experiment was to determine whether adding lines to the bottom of the Jastrow illusion, thereby causing it to resemble the Òshorter lineÓ of the Oppel illusion, would aid, hinder, or have no effect on the original Jastrow illusion. Subjects performed the experiment three times, with alternating differences added into the original illusion in the second and third tests.
Methods
The subjects involved were 20 Alma College student volunteers, aged 18 to 22. Twelve females and eight males participated. The experiment was conducted in a psychology computer laboratory. The illusions were created using the program Eye Lines. First, we ran the original Jastrow illusion. A horizontal line was then added to the center of the second crescent, connecting the top and bottom arcs. In the third test, bars were added to the second crescent again, this time one on each side, made to appear as a continuum of the end bars on the first crescent. The second crescent lays under the first in all trials. (See Figure 1.)
Each of the participants sat in front of the screen and went through a series of nine trials - three for the original Jastrow, and three trials for each of the second and third tests. The stimuli were presented in a random order determined by the computer. Using the mouse, each participant adjusted the top crescent to make the top and bottom crescents appear equal in length. The results of each trial for each participant were transferred to Statview, where a mean and standard deviation were calculated for each.
Results
The graphical results of the three stimuli are apparent in Figure 2. There
did not appear to be a significant difference between the three types of
illusions. For Trial 1, the original Jastrow, the mean was 74.8 and the
standard deviation was 73.94. In Trial 2, the illusion with one
equidistant line, the mean was 78.2 and the standard deviation was
74.43. In Trial 3, the illusion with two lines added, the mean was 64.75
and the standard deviation was 97.21. These values are too close to one
another to indicate a significant change.
Discussion
The data and the graph indicate that adding lines to the Jastrow illusion plays a relatively small role in changing the effectiveness of the illusion. Adding one line to the illusion had no more of an effect than the original Jastrow. Adding two lines, while showing a slightly larger range of variability, still did not have a significant effect on the original Jastrow illusion. Because the two lines did show some change, however, an idea for future research might be to add greater than two lines to the bottom crescent. Additional lines might counteract the effect of the original illusion, or it may follow the trend of the other changes.
References
Erdfelder, Edgar & Faul, Franz. (1994) Class of information-integration models in the Oppel-Kundt illusion. Zeitschrift fuer Psychologie, 202. 133-160. From Psych Abstracts.
Murray, Frank B. (1968) Cognitive conflict and reversibility training in the acquisition of length conversion. Journal of Educational Psychology, 59. 82-87.
Pick, David F. & Pierce Kent A. (1993) Theoretical parallels between the Ponzo illusion and the Wundt-Jastrow illusion. Perceptual & Motor Skills,76. 491-198.
More Posters
Web Posters from other courses and projects
Alma College Psychology Department