Several studies have shown the importance of depth cues in understanding illusory contours such as the Kanizsa triangle (Petry and Meyer, 1987). It has also been shown that completion of broken figures is in part a time related task (Murray, Sekuler, and Bennett, 2001).
However, little research has been done to show the effects of time on depth perception, whether subjects are able to see illusory contours in very short time frames. There has also not been much research into whether participants in shape completion tasks can recognize in short time spans that they are seeing an incomplete shape and filling it in with their minds, of if they see the shape as whole when it is presented.
Methods
The subjects were ten undergraduate students at Alma College. Half of these students were men and half were women. They were not informed of the illusions that were being studied. They were each tested individually.
The study was done using a 17 inch monitor attached to a Macintosh computer equipped with the following software; EyeLines and Aldus Super Paint. The EyeLines software presented the stimuli and recorded data from each trial, as well as functioning as a tachistoscope. The stimuli were prepared and the graphs created using Super Paint.
The first set of images are shown in Figure 1. These images were flashed in ramdom order, three times each. Each image was flashed for .15 seconds, .25 seconds, and .35 seconds. Before each new image was flashed, a small dot appeared on the screen functioning as a focal point for the participants.
The second block of images contained the images in Figure 2. These images were flashed in ramdom order, three times each. Each image was flashed for .15 seconds, .25 seconds, and .35 seconds. Before each new image was flashed, a small dot appeared on the screen functioning as a focal point for the participants.
The images were flashed using the tachistoscopic method. The subjects were instructed to view the images and then answer a series of questions about the images. The questions for the first block were: What did you see? Was the image a pattern or a shape? Was any part of the image brighter than the screen? The questions that were asked for the second block were: What did you see? Was it a complete or broken image? If it was broken, what parts were missing?
Figure 1 The Kanizsa Triangle and the Rectangle Image were both shown separately to the subjects. Each image was shown for a selected period of time. The subjects were asked to describe what they saw.
Figure 2 The Broken and Completed Squares and the Broken and Complete Ps were flashed at random. The subjects differentiated between all of the objects and also filled in the broken pieces to make sense of the objects
Results
Of the subjects tested one male and one female observed a the shape of a rectangle when shown the Rectangle Image. Four males and four females observed a pattern of four lines and four dots when shown the Rectangle Image. None of the subjects identified any part of the screen as being brighter. See Figure 3. Upon seeing the Kanizsa Triangle illusion, one female identified a triangle formed by dots in its corners. Four males and four females identified a triangle formed by angles at it's points but missing the parts of the connecting lines. One male described seeing a pattern with no distinct shapes. None of the subjects identified any part of the screen as being brighter. See Figure 4. After the experiment was completed, the subjects were shown the images for an unlimited amount of time. The subjects were all able to see the illusions and identified the inside shapes as being brighter than the rest of the screen.
For the second block of images, all of the subjects correctly idnetified the completed suare and completed letter P. When the image of the incomplete square was shown, all of the subjects identified the image as a square with parts missing. When the image of the incomplete letter P was shown, all of the subjects identified the image as a P with missing parts. See Figures 5, 6, and 7. All of the subjects correctly identified where the missing portions were in each of these shapes. Each subject used the existing cues to recognize the broken images as simple common figures.
Figure 5. The Effects of Time on Completion Recognition at .15 Seconds.
All subjects were able to correctly identify completed or broken figures after viewing them for .15 seconds. All subjects were also able to easily identify where the figures had been broken.
Figure 6. The Effects of Time on Completion Recognition at .25 Seconds.
All subjects were able to correctly identify completed or broken figures after viewing them for .25 seconds. All subjects were also able to easily identify where the figures had been broken.
Figure 7. The Effects of Time on Completion Recognition at .35 Seconds.
All subjects were able to correctly identify completed or broken figures after viewing them for .35 seconds. All subjects were also able to easily identify where the figures had been broken.
Discussion
Through this experiment, we have found that the illusory effects of the images are not always apparent when the eye has only seen them for less than a second. The Kanizsa triangle in the first block of images had the most illusory effect. Some of the subjects were able to see a triangle formed by the lines of the corners and others were able to see a triange formed by the dots in the corners of the triangle. The rectangle illusion, however, did not produce as varied a response. Some subjects were able to recognize a rectangle formed by the dots in the center. More subjects only saw a pattern of lines and dots and did not identify any shapes. This allows us to conclude that the Kanizsa and similar images will not have a strong illusory effect when briefly flashed at the eye. The eye does not have adequate time to process the image as much as it does when it is able to focus on it for longer. It has been shown that the illusion works due to the depth cues the subjects pick up. In shorten time periods, they can not process these depth cues. When the subjects were allowed to look at the images after the experiment, all of the subjects identified the illusion and notied the center images were brighter than the rest of the screen.
The second block of images allows us to conclude that subjects do fill in the missing parts of an image to make a recognizable image. Although the image was presented to them in less than a second, the subjects were able to recognize that there were places missing and they were albe to correctly identify where the missing pieces were. We can conclude from this that our hypothesis that viewing time affects the strength of the illusion is correct.
Further research should be done to determine the smallest amount of time necessary for seeing the illusion. This could be done by showing Kanizsa and Kanizsa type illusions for varying lengths of time (in a time sequential order) and having participants note when they can see the illusion.
References
Murray, R.F., Sekuler, A.B., & Bennett, P.J. (2001) Time Course of Amodal Completion Revealed by a Shape Discrimination Task. Psychonomic Bulletin and Review, 8, Part 4, 713-720.
Petry, Susan, & Meyer, Glenn E. The Perception of Illusory Contours. 1987. Springer-Verlag New York Inc. New York, New York.