Abstract
Does the same basic-level advantage commonly observed in the categorization literature also hold for targets in a search task? To answer this question, we first conducted a category verification task to define a set of categories showing a standard basic-level advantage; for correct true trials the basic level was categorized the fastest (612ms), followed by subordinate (642ms) and superordinate (702ms) levels. We then created a search experiment using the same target categories and photorealistic objects, now arranged into six-item search arrays. Participants were cued with a category name at one of the three levels, and then shown a target-present/absent search display. Target-present displays included one target among five distractors drawn from different non-target superordinate categories. Target-absent trials contained five such distractors along with a single lure at the same hierarchical level as the target. For target-present trials, we found that the time between search display onset and fixation on the target, our measure of search guidance, was shortest for subordinate cues (426ms), longer for basic-level cues (462ms), and longer still for superordinate cues (530ms). However, target verification, the time between target fixation and the button press, showed the standard basic-level advantage; basic (432ms) was faster than subordinate (492ms) and superordinate (522ms). For target-absent trials, gaze generally took longer to reach the lure, and this resulted in the basic-level advantage appearing in guidance (basic, 518ms; subordinate, 549ms; superordinate, 675ms), but not in verification (no reliable differences between levels). These findings demonstrate different hierarchical advantages for guidance and verification in categorical search. Search guidance shows a subordinate-level advantage; the more accurately a target template can be described, the better the guidance. Verification shows a standard basic-level advantage. Caution is therefore required when interpreting manual categorical search times, as they reflect a mixture of different and partially offsetting guidance and verification processes.
Meeting abstract presented at VSS 2012