Eliminating Depth Perception Weakens Ponzo Illusion

The Ponzo Illusion was first demonstrated by Mario Ponzo in 1913 and has often been attributed to depth perception. The illusion contains two horizontally parallel lines that lie between two vertically parallel lines which are acutely slanted towards each other. The illusion causes the top line to appear longer than the bottom line, though they are the same length. Theories to account for this illusion include the Confusion Theory, Tilt Constancy Theory, Linear Perspective Theory, Orientation Theory, and Depth Perception Theory. The Depth Perception Theory suggests that images that seem far away will appear larger than actual size due to the brains tendency to overestimate the size of distant objects. The Depth Perception Theory suggests this to be the rationale for the Ponzo Illusion.

Newman and Newman (1974) illustrated that images suggesting depth in the Ponzo Illusion influence the strength of the illusion, while Fisher (1968) utilized the edges of a pointed roof that projected towards the observer as converging line elements. These experimenters found correlations between depth and the strength of the illusion, a finding that supports the Perspective Theory.

Most recently, Prinzmetal, Shimamura, and Mikolinski (2001) conducted several similar experiments that ultimately suggested that the genesis of the illusion is the misperception of orientation induced by visual cues. Their findings supported the hypothesis that the Tilt Constancy Theory better explains the Ponzo Illusion than the perceived depth factor.

The intention of this experiment is to conduct a test that offers contending insight of Prinzmetal, Shimamura, and Mikolinski latter findings. The aim is to decrease depth perception in the Ponzo Illusion to better determine the validity of depth influence. The hypothesis is that the elimination of depth perception in the Ponzo Illusion will decrease the strength of the illusion, in accordance with the Perspective Theory.

To test this hypothesis, a stimulus that eliminates depth perception was presented, along with a control of two horizontally parallel lines, and the actual Ponzo Illusion. This stimulus is a picture of an object (an ice cream cone) that has virtually no depth, as compared to the railroad image with which the Ponzo Illusion is often associated. This experiment is not intended to disprove these theories, but to strengthen the validity of, and gain insight into, the Perspective Theory in terms of the influence of depth perception in the Ponzo Illusion.

Methods

Participants
Participants were 25 college students between the ages of 18-22, including 13 males and 12 females. The students were randomly selected for this experiment.

Apparatus and Stimuli
The experiment was run using Eye Lines software (Beagley, 1991), on a Macintosh computer, where the stimuli were displayed on a 20-inch screen. The visual images on the screen were constructed by the three group members of the experiment. Stimuli consisted of the Ponzo Illusion [Figure 1], a picture of an upside-down ice cream cone [Figure 2] which was scaled to match measurements of the original Ponzo and consisted of converging lines which diminished the depth. A control [Figure 3] which consisted of two parallel lines running horizontally was also used to properly compare error.

Procedure
Each participant sat down at the computer and the procedure was explained. The participants keyed into the program by typing their initials and then pressing enter in order for the stimuli to appear on the screen. Participants then adjusted the bottom line of each trial using the computer mouse to alter the line so that it appeared the same size as the top. This was repeated a total of 9 times to obtain 3 measurements on each of the stimuli. Subjects cycled through all stimuli randomly.

Figure 1

Figure 2

Figure 3

Results

Analysis of the data shows that participants had lower error when observing the Cone stimuli than when observing the Ponzo Illusion. As shown in Graph 1, the Control stimuli yielded the least mean error of the three types of stimuli, with a mean error of .592 mm. The Ponzo stimuli yielded slightly more mean error (1.091 mm) than the Cone stimuli (.862mm).

Graph 1

References

Fisher, G.H. (1968). An experimental comparison of rectilinear and curvilinear illusions.British Journal of Psychology, 59 , 23-28.

Newman, C.V. & Newman, B.M. (1974) The Ponzo illusion in pictures with and without suggested depth. American Journals of Psychology, 87, 511-516.

Prinzmetal, W., Shimamura, A.P., & Mikolinski, M. (2001). The Ponzo illusion and the perception of orientation. Perception & Psychophysics. 63 (1). 99-144.

More Posters

Web Posters from other courses and projects

Alma College Psychology Department