It has been demonstrated by Liebert and Burk (1985) that subjects can voluntarily control their perception of ambiguous figures. After researching this, the following study was designed to test the effects of subliminal and conscious influences on the perception of reversible figures. Experimenters expected that the subliminal and conscious messages pertaining to one interpretation of an ambiguous figure would make subjects more likely to see that interpretation first.
Subjects included 13 male and 14 female college undergraduates, chosen randomly.
Three ambiguous figures sized 6 X 4 inches were chosen, all centered on pieces of 8 X 11 paper. The pictures included the man/girl, man/woman and baby, and swan/squirrel figures.
SoundEditª 16 Version 2 for Macintosh computers was used to create sound tracks, each between 20 and 30 seconds long. The tracks featured two different voices telling two different stories, that would feed one into each ear simultaneously through headphones. Examples of the topics for the irrelevant stories included peanut butter and jelly sandwiches, a yellow cat, rainbows, shopping, and the solar system. The story pertaining to the man/girl illusion was about an "Uncle Dan." The story about the man/woman and baby figure described the woman and baby figure. The story for the swan/squirrel illusion was about squirrels in the park.
Subjects were randomly assigned to one of three groups. Subjects were asked to listen only to the right ear and repeat this story outloud, a technique called dichotic listening (Cherry, 53). This prevented subjects from hearing the story in the left ear, so that it would feed subconsciously. All groups listened to two practice trials to learn dichotic listening. The Control Group heard story sessions of no relevance to the ambiguous figures. The Subliminal Group repeated irrelevant stories as the stories pertaining to an interpretation of the picture went into the left ear. The Conscious Group repeated the relevant stories while irrelevant went in the other ear. After each story session, subjects in all groups were shown an ambiguous figure and asked to describe the first visible perception.
RESULTSNo subject from any group saw the girl first in the man/girl figure and only one subject (from the Control Group) saw a woman version of man/woman and baby. Figure 1 demonstrates the data gathered from the Control Group. The swan/squirrel figure was observed six out of nine times in each the Subliminal and Control Groups. Figure 2 represents the results from the Subliminal Group. Eight subjects from the Conscious Group saw the squirrel version, while only one reported seeing the swan first in the swan/squirrel figure. The results from the Conscious Group are represented in Figure 3.
Figure 1 - Control Group. All subjects from the Control Group saw the man versions of the man/girl and man/woman and baby figures. Six subjects saw the swan while three saw the squirrel in the swan/squirrel figure.
Figure 2 - Subliminal Group. The results from the Subliminal Group are identical to those from the Control Group for every figure.
Figure 3 - Conscious Group. All the subjects saw the man version of the man/girl picture. One subject out of nine saw a woman version of the man/woman and baby figure. Eight of the nine subjects perceived a squirrel in the swan/squirrel ambiguous figure.
DISCUSSIONThe results from the Conscious Group for the swan/squirrel picture were the most significant in supporting the hypothesis. Because subjects in the Conscious Group were hearing and repeating a story about squirrels, they responded immediately, without hesitation, picking out the squirrel figure. These results are significant when compared to the results from the Control Group which suggest that the swan is a more common figure to initially observe.
Although all subjects from Conscious and Subliminal groups saw the man in the man/girl after hearing the story, these results do not necessarily support the hypothesis because all the subjects from the Control Group also saw the man first. This indicates that there is an inherent bias in the figure causing the man to be the more dominant perception. Almost all subjects saw the man first in the man/woman and baby picture, suggesting the same bias. This also disproves the hypothesis, since the relevant story described the woman and baby interpretation.
One possible explanation for why both of the men figures were observed first lies in the fact that both the men interpretations were faces, while those of the girl and woman and baby were whole or nearly whole figures. The eyes of the men were also the center focal point of the pictures. In contrast, the swan/squirrel figure portrayed no face characteristics and the center of the figure was no more relevant to one interpretation than the other.
In future studies, experimenters may want to avoid this bias, by making sure pictures are equally noticeable during initial observation. This could be done by running tests of the figures, much like the Control Group, to ensure both interpretations occur at similar frequencies. Figures should either both be faces, or whole figures.
REFRENCES
Liebert, Robert M. & Burk, Bonnie. (1985). Voluntary Control of Reversible Figures. Perceptual and Motor Skills, 61, 1307-1310.
Cherry, E. C. (1953). Some Experiments on the Recognition of Speech With One and With Two Ears. Journal of the Acoustical Society of America. as cited in Anderson, John R. Cognitive Psychology and Its Implcations 3rd ed. W. H. Freeman: New York. 1990