Prioritizing feature bindings across space and modality in working memory.

A growing body of evidence shows that selective attention can be strategically directed to prioritize items of higher "value" in working memory. This work has typically been limited to tests of feature binding using simple "unitized" colored shapes as memoranda. Recent research has suggested prioritization may not be effectively applied to color-odor bindings. This raises the possibility that the benefits of value-based prioritization in working memory may be limited to bindings between unitized visual features (e.g., object-color bindings, object-location bindings). The current study explored this, examining whether value-directed prioritization can be effectively applied when feature pairings of color and shape are separated out, either into different spatial locations (Experiment 1) or across visual and auditory presentation modalities (Experiment 2). Experiment 1 found an overall cost to working memory performance when features were spatially separated, relative to a unitized condition, while Experiments 2 found no such cost for cross-modal feature separation. Across the experiments, participants were equally able to derive performance benefits from prioritizing high-value items in the sequence, regardless of whether features were encountered as part of the same unitized object or separated in space (Experiment 1) or modality (Experiment 2). The findings have implications for the relationship between working memory and attention, suggesting that value-directed prioritization can be effectively applied across different types of feature binding.