Abstract
Attention allows cognition to focus on task-relevant items at the expense of others. While selection over visual input has been extensively studied, attention also operates on information in the mind. To what extent do the mechanisms and properties of mental selection mirror those of perceptual selection? One of the organizing principles of perceptual attention is that it is influenced by how individual features and parts are grouped as integrated objects. Here we ask whether selection in working memory is object-based by utilizing two influential paradigms in visual attention. In Experiment 1, we adapted the double-rectangle paradigm (Egly, Driver, & Rafal, 1994) into a memory updating task where participants memorized colored squares at the end of two different bars and simultaneously updated the colors of two target squares within memory. Updating targets on the same object is faster, F(1,16) = 6.78, p = .019, and more accurate, F(1,16) = 5.35, p = .034, than updating those on different objects even when the distance between targets are the same. In Experiment 2, we examined the spread of attention to task-irrelevant features in a retro-cue task. Participants memorized colored Gabor patches that were sequentially presented and received both an item and a feature dimension cue (color or orientation) with 70% validity. We found a main effect of cue validity, F(2,15) = 12.49 , p < .001. There are lower errors in an adjustment task for validly cued trials than both the invalid same-feature, p = .002, and invalid same-object trials, p = .016. Importantly, we found evidence for object selection over feature selection: Participants were more precise in the invalid same-object condition than in the invalid same-feature condition, p = .005. Together these findings show that selection in working memory is similar to visual attention in that it acts on coherent, bound objects.