The Attention Operating Characteristic (AOC), introduced by Norman and Bobrow (1975) and formalized by Sperling and Melchner (1978), is the attentional analog of the ROC curve. It plots performance on Task A (y-axis) against performance on Task B (x-axis) as the observer shifts attention between tasks. The shape of the AOC curve reveals the nature of the attentional resource being shared.
AOC Shapes and Their Interpretations
Linear: fixed resource, proportional sharing
Right-angle: independent resources (no tradeoff)
Point at upper-right: unlimited capacity (no cost)
A concave AOC indicates that the two tasks compete for a limited shared resource: improving performance on one task necessarily degrades the other. A rectangular AOC (performance on both tasks near ceiling except at the extremes) suggests independent processing resources. The slope of the AOC at any point gives the exchange rate — how much performance is lost on one task per unit gained on the other.
Applications
AOC analysis has been applied to dual-task paradigms in attention research, workload assessment in human factors, and studying the capacity of visual attention. It provides a more complete picture of attentional allocation than single-task performance measures, revealing whether divided attention costs are symmetric or asymmetric and whether they reflect data limitations or resource limitations.