Two components and two stages in search performance: a case study in visual search

A model of skilled performance in visual search tasks is presented. According to the model, search behavior is controlled by two memory components. Each of them plays a different role in two stages of search. The two components are the memory representations of the target category and context category, respectively. The two stages are scan and check-for-target. It is argued that the scan is dominantly context-controlled (resulting in target detection) and that the target check is target-controlled (leading to target identification). The model is tested in a transfer experiment. It is shown that the model must be revised to account for two different ways of processing, depending on properties of the stimulus lists: (i) one-stage-scan under target and context control and (ii) scan-and-check processing where checking is target-and-context controlled and scanning is not under memory control at all. Implications of these findings are discussed.