The Effects of Changes in Angle Size on the McCollough Illusion
Sensation and Perception
Alma College
Celeste McCollough (1965) first described the color aftereffects which are dependent on two kinds of stimuli: color, most often complementary colors such as green and magenta; and contour orientation, most often square-wave gratings presented horizontally and vertically.
The illusion is created by adapting the subject to gratings which contain the stimuli, both color and orientation. The subject then views the black and white test gratings which are oriented like the stimuli. At this time the subject generally experiences the McCollough aftereffect: the white sections of the test gratings appear to be tinted with desaturated color complementary to that of the stimuli.
McCollough's original study and further research (Sharpe & Tees, 1978; Humphrey & Goodale, 1998; McCollough, 2000) have indicated that the gratings need to be orthogonally related for the aftereffect to be produced. The aftereffect is dependent on the fatiguing of neural mechanisms that code for color as well as orientation. Fidell (1970), however, demonstrated that the illusion is still visible, although weakened, at angles less than 90 degrees.
In the experiment reported here we attempt to demonstrate that this illusion can be produced simultaneously with two pairs of complementary colors of gratings placed at right angles to each other. We also investigate Fidell's research further as we examine the effects of reducing the angle orientation from 90 degrees to 45 and finally to 0.
The subjects for this experiment consisted of two males and four females, including the two researchers. Their ages varied from 20 to 57. Other than the researchers, the subjects have has little or no experience with visual illusions.
Materials
Square wave gratings for both the stimuli and the test grating were created on a Macintosh computer using Adobe Illustrator and Adobe Photoshop. The gratings were converted to and viewed with Eye Lines (Beagley, 1997) software. The test grating was black and white (Fig. 1) and the stimuli were black with red, black with green, black with blue, and black with yellow (Fig. 2&3). The complementary colors (green/red; blue/yellow) were produced with Adobe Illustrator in saturated hues and were presented at right angles to each other in the first trial. In the second and third trials the red and green remained constant but the blue and yellow were presented at 45 degrees to each other and at 0 degrees to each other.
Each subject participated in a total of three trials which were a minimum of 24 hours apart to eliminate aftereffects from previous trials. In each trial the stimuli were presented with Eye Lines software on a Macintosh computer in a darkened room. The subjects viewed each grating successively for 20 seconds, then for 10 seconds and finally for 5 seconds. The subjects then viewed the test grating and reported color aftereffects and other visual illusions on the grating.
In all trials more than half of the subjects reported a red color aftereffect in grating D. (Fig. 1). Only in trial one were all color aftereffects reported. Trial two retained red, green and yellow aftereffects but blue was no longer seen. Finally in trial three only red and green aftereffects were present.
Of the six subjects tested in trial one, three reported seeing green aftereffects in grating A (Fig 4). Two subjects reported yellow aftereffects in grating B. Three subjects reported a bluish aftereffect and two reported a gray aftereffect in grating C. All subjects reported a red or pink aftereffect in grating D. Although some aftereffects were stronger than others, the effects that were reported were consistent with McCollough's original research and with Fidell's research using two sets of complementary colors.
Of the six subjects tested in trial two, one reported green aftereffects and two reported a grayish color in grating A. One subject reported a yellow aftereffect and two reported a grayish color in grating B. All subjects reported some color described as a shadowy hue in grating C. All subjects reported a pink or red color, although it appeared less saturated than the previous trial in grating D. As the angle of orientation of the blue and yellow stimulus gratings was reduced the aftereffects also seem to be lessened.
Of the six subjects tested in trial three, one reported a green aftereffect and two reported an undetermined color in grating A. No color aftereffects were reported in either gratings B or C. Four subjects saw a pinkish colored aftereffect in grating D. Consistent with Fidell's research as orientation of the blue and yellow angles was eliminated so were the color aftereffects in the test gratings (B&C).
Trial one supported our original supposition (and Fidell's research) by demonstrating that four color aftereffects can be simultaneously produced. The results were not as homogeneous as expected: the red or pink aftereffect appeared more pronounced than the others. As the blue and yellow gratings were presented in less orthogonal orientations the aftereffect diminished and finally disappeared as we expected. However, it was apparent from our results that the reduction of aftereffect in the last two trials was also due to the discrepancy in the angles of the test grating. The angles for gratings B and C should have been changed to match the blue and yellow stimuli. To determine if these results were caused by this discrepancy will take further experimentation.
Two other aspects of this experiment also raised questions. First, many of our subjects reported seeing "discolorations" or effects of undetermined color. We are uncertain whether these artifacts were color aftereffects that were hard to discern or possibly lateral inhibition from the black grid lines. Further research would be needed to determine the exact origin of these visual illusions. Although we reported these observations in our results we did not graph them (Fig. 4), as we considered only reports of complementary colors to be truly representative of the McCollough illusion.
Another outcome of this research that had not been considered but seemed to affect our subject's responses was that of suggestibility. The researchers reported stronger color aftereffects than did the subjects that were unfamiliar with the illusion and its expected results. This indicates that this illusion can in part be created by previous expectations. Further research in this same area would be more accurate with inexperienced subjects. Investigating the effects of suggestibility role in this illusion would be warranted.