Directed avoidance and its effect on visual working memory

Attentional control processes help to prioritize the storage of information in visual working memory (VWM) by gating what enters the system and influencing how precisely this information is stored. However, the extent to which such prioritization occurs deliberately, opposed to incidentally, is poorly understood. In large part, this is because investigations of this matter have almost exclusively relied on comparisons of memory for exogenously cued items versus uncued items. To understand whether prioritization occurs independent of intention, though, it is essential to examine instances in which attended items are entirely task-irrelevant. Thus, in the current study we used a directed avoidance paradigm to examine VWM performance following the selection of an item known to be task-irrelevant. In Experiment 1, we confirmed that cueing the color of a non-target item paradoxically increases attention to the cued item when the target color is unknown, resulting in longer search times (in line with previous findings). In Experiments 2 and 3, we applied the same cueing procedure to a delayed-estimation task of VWM, but now found a non-target cueing benefit in which the recall of task-relevant items was improved by directed avoidance. We further found that this effect is not solely due to the reprioritization of cognitive resources during maintenance (Exp. 4), but involves additional control processes that 1) reallocate resources to relevant items at encoding, and 2) selectively stabilize such items during the transition from encoding to maintenance (Exp. 5). As such, we suggest that while attentionally selected items may initially be prioritized independent of importance, more controlled mechanisms reallocate resources on the basis of relevance when sufficient time is provided before the sensory information is removed or displaced.