Dual-Target Cost in Visual Search for Multiple Unfamiliar Faces
23.3032, Saturday, 16-May, 8:30 am - 12:30 pm, Banyan Breezeway
Natalie Mestry1, Tamaryn Menneer1, Hayward Godwin1, Kyle Cave2, Nick Donnelly1; 1Psychology, University of Southampton, UK, 2Psychology, University of Massachusetts Amherst, MA, USA
Visual search is slower and less accurate when trying to find more than one target object, referred to as the dual-target cost, and reduced attentional guidance has been shown to be the cause (e.g. Menneer et al., 2012). Studies in the eye-witness domain have shown the presence of two target faces reduces accuracy of identifying a single target face in both memory and matching paradigms (Megreya & Burton, 2006; Bindemann et al., 2012; Megreya & Bindemann, 2012). Here we examine behavioural performance and eye movements when searching for one versus two unfamiliar faces. Specifically, we were interested to see whether there was a dual-target cost and, if observed, whether the cause is the same for faces as for objects. Across two experiments, we varied the visual similarity of distractors to target faces using morphing. In both experiments, the similarity of distractors to targets was graded. In Experiment 1 distractors were morphed with only one of the targets and in Experiment 2 some distractors were morphs of both targets. In both experiments there was evidence of a dual-target cost: dual-target search took longer and was less accurate than single-target search. On target-absent trials, the participants searched exhaustively in both single and dual-target conditions, highlighting the difficulty of the task. Searching for two targets resulted in ‘shedding’, wherein participants gave up searching for one of the targets. There was evidence of guidance to faces in relation to the participant’s preferred target in both single and dual-target search in Experiment 2, but only single-target search in Experiment 1. Probability of fixation to distractors was proportional to the similarity of the distractor to the preferred target. Overall, results suggest effective visual search for unfamiliar faces is limited to a single face, which has profound implications for applied search tasks.