Abstract
False Pop Out occurs when a distractor(s) in a visual search display poses as a target by drawing responses away from the actual target. It has been observed in classic singleton displays with homogeneous distractors and in heterogeneous displays where basic features differences would have led classic visual search theories (FIT, GS) to predict otherwise. It has also been observed in a pure form: in 3-item displays wherein one item is metameric with the second but anti-metameric with the third. This results in one of the distractors (i.e., one of the two identical items in the display) being perceived as the unique target almost 100% of the time. While FPO has been observed in displays varying in size and stimulus composition, the question remains as to whether these displays can produce the RT functions (specifically flat RT slopes as a function of set size) indicative of classic pop out. The results from a controlled experiment specifically looking at RT functions for FPO will be presented as converging evidence for our hypothesis that pop out does not result from basic feature differences in displays, but rather is the result of broken emergent features (specifically, symmetry) in displays that contain inter-stimulus configural relationships (i.e., displays processed as Gestalts). This hypothesis is born from the Theory of Basic Gestalts (Pomerantz and Portillo, 2011), which has demonstrated that when parts configure, unique-item search can be as efficient as differentiating black from white (Pomerantz, Sager & Stoever, 1977), and that these configural superiority effects remain stable with increasing display set sizes. If configurations can reliably improve unique target search, they should also reliably provide for FPO across display sizes. That is, items that break symmetry will truly pop out, regardless of display set size and regardless of whether they are unique.
Meeting abstract presented at VSS 2014