Abstract: Inhibition of return (IOR) refers to a bias against returning attention to a location that has been recently attended. Previous study has demonstrated that spatial working memory (WM) play a key role in determining and maintaining inhibition at cued locations. However, Tipper et al (1991) found that IOR not only existed in static but also in dynamic displays. It thus remains unclear which type of WM underlies the maintenance and updating of IOR in dynamic displays. Moreover, Despite IOR in dynamic displays has received a great deal of research, no consensus has yet been reached regarding the nature of the IOR in dynamic displays: Tipper et al (1991) suggested that IOR in dynamic displays was inhibited from return to recently attended objects. Robertson et al (2001, 2004), however, suggested that this effect could be reinterpreted as spatial reorienting processes. The present study aimed to examine 1) which type of WM underlies the maintenance and updating of inhibition in dynamic displays and 2) whether object-based IOR in dynamic displays was resulted from the updating of the spatial inhibition or the inhibition of object representation.
Method
The study employed Three 2×2 factorial within subjects design. The two factors are cuing (the target presented on the cued object, the target presented on the un-cued object) and have WM load or not (have WM load, have no WM load), with voice WM, spatial WM and visual WM load in three experiment respectively. 60 undergraduate students form northeast normal university volunteered to participate for paid. They ranged in ages from19 to 23 years old. All participants were naive to the purpose of the experiment and had normal or corrected-to-normal vision. Stimulus presentation and data collection were controlled by a LENOVO microcomputer running E-prime software. By using the dual-task paradigm, three experiments investigated the role of the voice, spatial and visual working memory in IOR, respectively.
Results
The results of present study showed that when the working memory task was non-spatial in nature (voice or visual), IOR was present, although overall reaction times were greater in the presence of the WM task. When the task involved a spatial WM load, IOR was eliminated.
Conclusions
Overall, the results suggested that spatial working memory system plays a key role in the maintenance and updating of the IOR in dynamic displays and provided new evidence for the idea that the so-called object-based IOR in the dynamic displays was the result of update of the spatial inhibition rather than the object inhibition

Key words: working memory, object-based inhibition of return, dual-task paradigm