Perceived Rather Than Physical Direction of the Double-Drift Stimulus Pops out in Visual Search

In the double-drift stimulus, a patch of 1/f noise moves in one direction while its internal texture drifts in the orthogonal direction and the result is that the perceived direction deviates dramatically from the physical path. Surprisingly, non-delayed saccades to the double-drift targets go to their physical not perceived locations (e.g. Lisi & Cavanagh, 2015). Using an oddball search paradigm, we examined whether visual attention operates over the perceived or physical direction of the double-drift stimulus. We presented either 4 or 8 moving patches, one of which differed in only its illusory direction from the others (each half of the items had matching physical directions) or only its physical direction from the others (all perceived directions were matched). Participants pressed a key once they localized the target, then another key to report target location. Results showed that the odd illusory direction produced a pop-out effect: the participants were able to report the target’s location correctly on 96% of the trials, independently of set size. Reaction times also showed that the target popped out among the distractors: the response to the odd target was faster in the larger set (4 items, 1.43 s vs 8 items, 1.10 s). This improvement with increasing distractor set size is consistent with previous findings of the pop-out effect in visual search (e.g. Bravo & Nakayama, 1992). In contrast, when the odd target was defined by its physical direction, participants showed poorer performance and even worse performance at the larger set size (40% and 27% correct, respectively). Average reaction time was, overall, longer and increased with set size (3.18 s vs 4.18 s, respectively). We conclude that pop-out operates over perceived, not physical double-drift directions. Perceived double-drift directions are subject to parallel processing whereas physical double-drift directions require serial search.