Interference control theory: A new perspective on dual-task interference in memorizing and responding to visual targets

Abstract

In a recent study, Nieuwenstein and Wyble (2014) showed that the consolidation of a masked visual target can be disrupted for up to one second by a trailing 2-alternative forced choice task. Aside from demonstrating that working memory consolidation involves a time-consuming process that continues after a mask, the results of Nieuwenstein and Wyble are remarkable in demonstrating retroactive interference (RI) with little to no proactive interference (RI) – the opposite of what is typically found in studies on the attentional blink and psychological refractory period effect. Here, we show that the reversal from PI to RI depends on the probability of T2 presence (p[T2]), such that a high p(T2) results in strong PI with little RI whereas a low p(T2) results in strong RI with little PI. To explain these findings, we propose that the degree of PI and RI reflects the workings of an attentional control mechanism that aims to protect T1 consolidation against interference, and that is applied in accordance with the risk of such interference. In this view, a high p(T2) entails a high risk of interference with T1 processing, and this results in the protection of T1 at the expense of a postponement of T2 processing. Converely, a low p(T2) means that the risk of interference is low and this entails that T1 is left vulnerable, while T2 can be processed unabated. Consistent with this account, we show that if the risk of interference with T1 processing is increased by embedding the targets in an RSVP stream of distractors, we find strong PI with little RI even when p(T2) is low. Aside from offering a new perspective on dual-task interference, this work makes clear that attentional control plays a central role in phenomena observed in studies using RSVP.