Abstract
Consider a paradigm in which a visual stimulus is first presented and then a feature of that stimulus appears as either a target or distractor in a subsequent visual search task. It has been shown that if a response is made to the initial appearance of the stimulus, a validity effect is found; search times are faster when the stimulus is part of a target than when it is part of a distractor. This validity effect disappears if a response is withheld to the initial stimulus. Our study demonstrates that there is a validity effect and an inverse validity effect. First, like previous studies, we replicated the validity effect when the first stimulus was responded to. Second, unlike previous studies, when the first stimulus was not responded to, in four experiments we consistently observed an inverse validity effect such that faster search times occurred when the initial stimulus was contained in a distractor. Third, when we changed the second task from a visual search to stimulus presented in isolation, only the inverse validity effect was found. Fourth, when we increased the overlap between the first and second response buttons, the inverse validity effect increased. Based on our findings, we argue that the validity effect is driven by biased competition; responding to the first stimulus increases the attentional weights assigned to that stimulus's features such that it wins the competition for selection in the search phase. The inverse validity effect, however, is driven by feature binding into event files as there is a partial repetition of the event file formed from the first response when responding to the search event.
Meeting abstract presented at VSS 2016