Attending to certain visual features or objects may have significant consequences for performance in some circumstances, but is not necessarily important in all tasks. Understanding the role of attention in the performance of visual tasks requires the characterization of the mechanisms of attention and a taxonomy of task conditions under which attention is important. We extend a powerful method and model for identifying and characterizing the effect of attention on perceptual performance. The method combines an external noise approach (adding visual noise similar to random TV noise) with different visual tasks similar to everyday or operator tasks involving complex displays to identify the mechanism(s) of attention in each task and to understand when attention is important to high performance. The approach provides a multi-dimensional characterization of task performance in terms of signal strength, external noise, and target similarity. We use a number of task manipulations to assay the role of attention, including spatial cuing, inhibition of return, visual search and visual short-term memory. The results from these empirical observations will be used to construct and test a taxonomy of visual attention. The goal is to develop an overarching system and theoretical structure to organize all the empirical observations about when and how and why attention can improve human performance. In addition, we evaluate whether and when training can eliminate attention demands.
Understanding when attention is a limiting factor in performance and whether training can eliminate attention demands will have theoretical and practical implications for understanding attention deficits in individuals exhibiting abnormalities of attention associated with mental health conditions. This analysis can additionally constrain process models of attention in normal observers and may suggest constraints on the relevant properties of neurological models of attention.
Showing the most recent 10 out of 11 publications
