Abstract
According to Target-Contrast Signal Theory (TCST), parallel visual search unfolds via the active comparison of a search template held in mind and all stimuli in the search array. The aim of the comparison is to reject non-target items. Attention then scrutinizes the remaining non-rejected items. Previous studies on TCST focused on visual search tasks that had a fixed template throughout the experiment. Here, we studied how peripheral parallel search unfolds with varying templates. Following research by Carlisle, Arita, Pardo and Woodman (2011), we examined search with a target template that changed every five trials (a sequence). The targets were either a red triangle or a cyan semicircle. Target-distractor similarity was varied with two different kinds of distractors: orange diamonds or blue circles. Only one type of distractor was present on each trial. In addition, set size was varied to dissociate search processes (search slopes) and non-search (overall RT) processes. Consistent with TCST, results showed that RTs increased as a logarithmic function of set size, with larger logarithmic slopes in the high target-distractor similarity conditions than in the low target-distractor similarity conditions. Importantly, search efficiency varied as a function of position in the trial sequence. Regardless of target-distractor similarity, search slopes were steeper earlier in the trial sequence, suggesting that evidence accumulation rates were slower. In addition, response times for the target-only conditions were not affected by trial position, suggesting that switching target templates only impacted search related processes. Our findings extend Carlisle et al.’s results by showing that only search specific processes are impacted by the switching of target templates between sequences of trials in peripheral parallel efficient search.