CHAMPAIGN, Ill. -- When there's a crisis in the cockpit, why do some flight crews think on their feet and react swiftly, while other crews make potentially fatal mistakes?
The question long has preoccupied airlines where passenger safety can rest on the capacity of pilots to cope with bad weather and equipment failures at the same time. Airlines have trained crews to follow three procedures with special care during an emergency -- collecting information about the situation, discussing the importance of the tasks and distributing the tasks among the members.
Research by a University of Illinois professor, however, suggests that an important element is missing -- a recognition that the timing of crew communications greatly affects performance.
"The conventional wisdom is 'more is better,' meaning that the more a crew engages in communications in an emergency, the better their performance," said Mary J. Waller, a professor of business administration. "What I found was quite surprising -- that simply discussing and distributing tasks across crew members was not associated with good performance."
Waller based her findings on a unique "micro" study of 10 crews that were similar in experience and training. The three-person crews (all white males) worked for the same airline and were videotaped in a sophisticated B-727 flight simulator.
Each crew "flew" the same pattern and faced a battery of problems, including a hydraulic system failure, bad weather and the loss of nose-wheel steering.
Waller coded crew behaviors at 10-second intervals after they were notified that weather conditions prevented them from landing at the scheduled airport. From then on, the crews were under uniform levels of time pressure, workload and rapidly changing conditions.
Waller used three senior commercial pilots to rank crew performances. She then cross-checked the performances with the type and quantity of conversations held during the emergencies.
"I tried to capture actual behavior as opposed to the more theoretical procedures often used in evaluating crew performances," Waller said. "My aim was the tear-apart patterns of behavior in groups on a micro level."
She found that crews that made mistakes had the same number of conversations as the high-performing crews, but did not engage in information exchange at the right time. "While
high-performing crews were very targeted and specific when an emergency arose, the low-performing crews tended to sprinkle their exchanges over the whole simulation. This amounted to a big disconnect between training and actual conditions.
"My research suggests that airlines -- or any organization where safety relies on team performance -- consider the issue of 'behavior timing' as a crucial element in the training of crews," the U. of I. researcher said.