Reliability and generality of the novel pop-out effect following passive pre-exposure to array items

Abstract

AbstractNovel pop-out refers to the relative ease in locating an unfamiliar target against a background of familiar distractors in visual search tasks. For instance, when one novel item is presented along with three familiar items, it is located faster than when the target is itself familiar, or when all items are novel or familiar. However, the reliability and generality of the novel pop-out effect has been questioned. Three experiments with human participants examined the reliability of within-array novel pop-out effects, and explored the conditions under which these effects can be obtained. The degree to which subjects could identify a novel item when presented against a background of familiar items, or against a background of other novel items, was assessed using two different visual search tasks. Experiments 1 and 3 replicated the theoretically important within-array novel pop-out effect, using the original procedure adopted by Johnston and colleagues. In the current studies, the array items were rendered novel or familiar by virtue of their previous absence or presence, respectively, in a pre-exposure phase, which allowed greater control over their novelty or familiarity. However, the use of this procedure abolished the typically-reported advantage for all-familiar versus all-novel arrays (Experiments 2 and 3). It is suggested that any model of attention should be sensitive to the present goal state of the subject, and that the literature on latent inhibition may provide one such mechanism through which such a set of effects could be explained.