Abstract: Correct readback and hearback are crucial to communication between pilots and air traffic controllers. Hearback errors can result in serious breaches in flight safety. 5351 controller-to-pilot transmissions were collected from an approach control radar simulator of the Civil Aviation University of China and transcribed. An analysis based on Logistic regression model was carried out to explore main factors resulting in controller hearback errors from three aspects of talk character, air traffic instance and individual character. The results indicate controllers are more prone to making hearback errors with the increasing departure and arrival aircraft, that traffic complexity has a negative significant impact on the possibility of hearback error, which means the more the traffic complexity is the less the hearback errors will be, and that a different ATS unit also has a significant impact on hearback error.

Key words: approach control, air traffic controller, hearback error, Logistic regression model, factors