Mohammad Shams^1,2 (), Peter Kohler^2,3, Patrick Cavanagh^1,2; ¹Department of Psychology, Glendon College, York University, Toronto, ON, Canada, ²Centre for Vision Research, York University, Toronto, ON, Canada, ³Department of Psychology, York University, Toronto, ON, Canada

A moving bar shifts the perceived position of a nearby flash. This effect is strongly asymmetrical: a flash at or ahead of the bar is pushed further ahead in the direction of motion but a flash behind the bar shows little or no shift in the opposite direction (Shams-Ahmar et al., ECVP 2022). Recently, it has been proposed that attentional repulsion (Suzuki & Cavanagh, JEP:HPP 1997) may cause the shift (Shams et al., ECVP 2023) as attention leads a moving bar (Szinte et al., J Neurophysiol. 2014). Here, we manipulate the expected motion direction to see if an attentional manipulation influences the motion-induced position shift. A shape moved downward to the center of the screen, where the dot flashed on top of it and the shape then moved away either rightward or leftward. In three sessions run on separate days, the probability of the second motion direction was left and right equally often, left more likely, or right more likely. We found that the illusory shift of the flash was consistently in the direction of the motion that followed it, and that the likelihood of the direction significantly affected the induced position shift (Friedman test: p=0.001; medians: unlikely=0.07 dva; ‘equally likely’=0.11 dva; likely=0.23 dva). Further, within each session, despite participants being aware of the more likely direction, the bias towards the more likely direction increased systematically across trials consistent with the notion of a gradual drift of attentional resources towards the expected direction of motion. No trial-to-trial effects were found, ruling out any contribution of serial dependence. We attribute the effects of direction frequency to a greater allocation of attention to the more frequent direction, increasing its effect on the shift. Attentional repulsion remains a viable explanation for the motion-induced shifts seen in static flashed tests.