A Simulator Study of the Effect of Information Load and Datalink on Crew Error

This study investigated the effects of air traffic control (ATC) induced information load and ATC communication modality on crew error in a fixed-base Boeing 747-400 flight deck simulator. To accomplish this, data from an experiment conducted by Boeing (Logan, et al., 1992) were re-analyzed. In this experiment, flightcrews engaged in two high-fidelity, full-mission scenarios. For this re-analysis, the variability in the amount of information ATC transmitted to the crew in the scheduled clearances was examined as an independent variable. To quantify this measure of information load, each scheduled ATC clearance was decomposed into mutually exclusive informational units called propositions. Each proposition consisted of one discrete “chunk” of information. On average, ATC clearances in the high information load condition contained more than two and a half times as many propositions as clearances in the low information load condition. Flightcrews made significantly more errors, and more errors per proposition, in the high information load condition than in the low information load condition. This indicates that it was the density of the information, or how many propositions were packed into each clearance, and not just the quantity of information which affected crew performance. Another manipulated variable was ATC communication modality. Each flightcrew participated in one scenario using a standard voice communications modality exclusively and in the other scenario using a combination of voice and digital DataLink. Though flightcrew error was significantly lower in the DataLink condition than the voice condition, this tendency was moderated by information load. In the low information load condition, flightcrews committed approximately the same number of errors in the voice and DataLink conditions, but in the high information load condition, flightcrews committed many more errors in the voice condition than the DataLink condition.