Purchase this article with an account.
Paul Dassonville, Walter Elizabeth; Roelofs effect demonstrates a ‘predictive’ use of unpredictable contextual location cues. Journal of Vision 2004;4(8):376. doi: https://doi.org/10.1167/4.8.376.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
When a visual target is presented within the context of a large frame whose center is located left or right of the observer's midline, the perceived location of the target is biased in the opposite direction: for example, a frame to the left of the observer's midline will cause an enclosed target to be perceived further to the right (the induced Roelofs effect, Bridgeman et al. 1997). Previously, we demonstrated that this illusion is caused by a distortion of the observer's egocentric representation of space, with the subjective midline pulled toward the center of the frame (Dassonville & Bala, VSS 2002). The current investigation sought to characterize the time course over which the context of the frame causes this distortion. Sitting in otherwise complete darkness, observers were asked to report the locations of brief targets (16 ms) presented just before or after the onset of a large frame. The onset time (750–1250 ms from trial start) and location of the frame (centered −6.6, 0 or +6.6 deg from objective midline) were randomly varied from trial to trial to prevent the observer from anticipating its characteristics. Surprisingly, the effect of the frame was seen even for targets presented as much as 66–100 ms before frame onset, as if the visual system were predicting the unpredictable location of the upcoming frame. These findings suggest that the frame's distortion of egocentric space occurs with a latency shorter than the time required to determine the location of the target. If this explanation of the effect is true, one would expect dim targets to be affected by the context of the upcoming frame even earlier than are bright targets, due to the longer neural processing delays for dim visual images. Indeed, the time course of the contextual effect was shifted approximately 12 ms earlier for the dim targets, providing further evidence that this ‘predictive’ effect can be attributed to a mismatch in the processing delays of the frame and target.
This PDF is available to Subscribers Only