Decrements in performance on a primary task associated with the reaction to voice warning messages

Voice warning messages are being adopted as an alternative to audio warnings because they can be readily understood, generate fast reaction times and supposedly allow a smooth transition from message to action. This experiment evaluated the effectiveness of synthesized voice messages using measures of performance decrement and response time in the context of a central warning system with audio, voice and visual indicators. The subject's task was to maintain accuracy on a psychomotor tracking task whilst responding to warnings. The results show that the various combinations of warning types could not be differentiated with respect to performance decrement on the primary task. The data for reaction time show that audio warnings produce the fastest responses, followed by voice warnings, with CWP indicators producing the slowest responses. The implications of the results for the role of synthesized voice warning messages in central warning systems is discussed.