Abstract
Does visual search freeze up during the central stage of information processing, at which the processes of working memory encoding and response selection take place? To address this issue, in the first experiment, we had participants identify and maintain a target (T1) embedded in a rapid serial visual presentation (RSVP) of digit distractors. The centrally presented RSVP was followed by a visual search task. Importantly, contrary to previous studies, a conjunction search task of searching for a green ‘T’ among green ‘O’ and red ‘T’ was used. We manipulated the SOA between the T1 and T2 (200-ms vs. 1200-ms) and set-size of visual search stimuli (3 vs. 6). The results showed that the T2 RT was significantly greater at the short SOA than at the long SOA, p < .001 and the T2 RT was significantly greater for the set-size 6 trials than for the set-size 3 trials across the SOAs, p < .001. Importantly, at the short SOA, the magnitude of set-size effect was significantly smaller than at the long SOA (23-ms vs. 47-ms), p < .005, revealing that increased processing duration of T2 was absorbed into slack time evoked by the concurrent process of working memory encoding. We suggest that this underadditive interaction between the SOA and set-size is because visual search proceeds concurrently during the attentional blink. In the second experiment, the first task was replaced to a four-alternative, speeded response task, followed by the same visual search task as in Experiment 1. The results of Experiment 2 also revealed a similar pattern of significant interaction between the SOA and set-size, p < .001, suggesting that visual search can also proceed during response selection stage. Taken together, we conclude that the perceptual processing of conjunction search stimuli took place during the central stage of information processing.