Relationship of Sector Activity and Sector Complexity to Air Traffic Controller Taskload

This study compared the relative effectiveness of two constructs, sector activity and sector complexity, in predicting air traffic controller taskload. Sector activity was defined as the activity associated with aircraft moving through the sector and was measured by counting the number of aircraft under the control of the sector during a traffic sample. Sector complexity describes a set of factors presumed to affect the difficulty experienced by a controller when controlling traffic. Sector complexity was measured in two ways. The first measure of complexity was a subjective rating made by supervisors and controllers to describe the complexity associated with specific traffic samples. The second was a composite variable that included measures reflecting several of the complexity variables found in the literature. Taskload was defined as controller activity and was measured by counting the number of data entries made by a controller during a traffic sample. The results appear to suggest that the authors' hypothesis, that sector activity predicted controller taskload better than sector complexity, was incorrect. However, interpretation of these results depended on consideration of what each of the variables measured. The Complexity Rating predicted controller activity better than the number of aircraft alone, but the Complexity Value (based on a set of variables identified through previous research) did not contribute at all to that prediction. Additional analyses suggested that the Complexity Rating measured something very different than the Complexity Value. The authors believe that the Complexity Ratings estimated the workload that observers believed the controller at the sector experienced instead of the complexity of the situation. On the other hand, the complexity measures used here did not appear to be not good measures of the construct. This may have occurred because the measures used in this study had limited variability or because they were not very good measures of the construct even though they were derived from factors identified in the literature as contributing to sector complexity. While the authors expected that the number of aircraft alone might be sufficient to predict controller activity/taskload, the results suggested that measuring both controller activity and extracting measures from other routinely recorded data might be necessary to develop more objective staffing standards used to determine how many controllers are needed to provide ATC services to individual facilities.