Visual working memory (WM) has a very limited online capacity making it considerably important to control the gating of encoding into WM. Recent studies have suggested that attention control is of importance in doing so, especially the time needed to disengage. However, the disengagement mechanism operates on a later stage of processing after the initial selection of information has already been initiated. We assume that individual differences in WM capacity are also driven by individual variations in the voluntary engagement of attention. In 2 experiments we investigated whether individuals with high- and low-WM capacity differ in the efficiency and speed of this attention control process. We realised different versions of the task in which different amounts of attention control were necessary, a more automatically triggered allocation of attention and a voluntary initiation of attention engagement, respectively. We further manipulated the time course to look for differences in the latency of attention control. The results revealed that participants with low-WM capacity were less effective to exhibit voluntary attention control processes and they were also slower in doing so compared with high-WM capacity individuals. However, this effect seems to be partly moderated by the ability to update the current task set. If the trial structure did not require task set updating smaller individual differences involving WM capacity could be found.

No abstract