Abstract:  Attention is biased towards the objects that sharing common features with the online representations in working memory (WM) and such phenomenon is called WM based attentional guidance. Although the attentional guidance was regarded as operating in an involuntary manner, many researchers still found that the guidance effects decreased or even eliminated with the increasing of WM load. However, the issue how WM load affects attentional guidance is still controversial. The probable reason for the attenuating of memory-driven guidance caused by the WM load might either be due to exhausting of cognitive resources with the load increasing, or be due to the reducing of the representation status when multiple representations maintaining in WM simultaneously. Here, with using the eye movement tracking technique to measure the attentional deployment in real time during the visual search task, we attempted to explore how the WM load modulated the attentional guidance under different perceptual load of visual search task. The classic dual-task paradigm combined the WM task and the visual search task was adopted in the present study. Participants were required to complete a visual search task while maintaining 1, 2 or 4 items in the WM online. During the visual search task, one of the WM items either reappear as a distractor of visual search task in the invalid condition or not reappear in the neutral condition. When low perceptual load of visual search task was used in experiment 1 and 3, the results showed that the distractor matched the WM representation could capture more of the first fixation than other distractors, and the RT in invalid condition was significant greater than the neutral condition in visual search under WM load of 1 and 2, suggesting a classic attentional guidance effect, but this effect was not observed when the WM load increased to 4 items which was regarded as the full load of WM. In contrast, when the high perceptual load of visual search task was used in experiment 2 and 4, the guidance effect was only found under the WM load of 1, and disappeared when the WM load add to 2 items. In conclusion, the present study illustrated that (1) both the load of WM and perceptual load of visual search task can modulate the WM based attentional guidance by the means of competing for the common cognitive resources, and (2) multiple representations maintained simultaneously in WM can guide attentional selection if the cognitive resources are sufficient.

Key words:  working memory load, perceptual load, guidance effects, cognitive resources