The contextual interference effect in visual feature binding: What does it say about the role of attention in binding?

Abstract

The effects of blocked versus mixed presentation were tested on visual feature binding, assuming that blocked presentation enhances focused attention, whilst mixed presentation recruits extra attentional resources for intratrial as well as intertrial processing. The contextual interference effect suggests that although performance due to mixed presentation is either similar or worse than blocked presentation when tested immediately, it is better when tested after an interval. We explored whether this robust empirical effect, common in psychomotor performance, would be evident in visual feature binding. Stimuli were conjunctions of shape, colour, and location. Study–test intervals from 0 to 2,500 ms were used with a swap detection task. In Experiments 1A and 1B, participants ignored locations to detect shape–colour bindings. In Experiments 2A and 2B, they ignored shapes to detect colour–location binding. In Experiments 3A and 3B, they ignored colours to detect shape–location bindings. Whilst Experiments 1A, 2A, and 3A used blocked presentation, Experiments 1B, 2B, and 3B used mixed presentation of study–test intervals. The results of these experiments and a replication experiment using a within-subjects design showed that the contextual interference effect appeared when spatial attention was engaged, but not when attention was object based.