Illusory Radial Displacement in Two-Flash Displays

While piloting a 2-alternative forced-choice experiment designed to quantify the visual system's precision for locating briefly flashed spots of light, I noticed that the second flash seemed closer to fixation, even when it wasn't. I measured the size of this illusion using the comparison-of-comparisons paradigm, in which observers were required todecide which of two pairs of flashes had the least difference in eccentricity. The average “wow factor” (i.e., bias divided by the standard deviation of performance-limiting noise) across observers was 0.5. There was no obvious relationship between illusion size and fixation behaviour. As yet, I have no good explanation for this “incoming illusion.”