Abstract
It was hypothesized that additional contextual linguistic redundancy in the wording of synthesized speech cockpit warnings would result in less required attention for initial comprehension and subsequent recognition. Twelve airline pilots were repeatedly presented 16 unfamiliar warnings in two wording formats, key word and sentence; e.g., Fuel low, compared to The fuel pressure is low. Then they heard the same warnings embedded in weather broadcast at a warning-to-weather sound pressure level ratio of +3 dB. For both listening conditions the task was to read back the message and also estimate the passage of time during message comprehension or recognition. Peformance on time estimation had previously been found to correlate with hypothesized attention demands during manual tracking and aircraft simulator flying tasks. As expected in the present study, intelligibility was higher and response time shorter for the sentence format than for the key word format. For unfamiliar messages, the tendency to underestimate the passage of time was greater for the two word format than for the sentence format, implying that less attention was available for the time estimation task during comprehension of messages with less linguistic redundancy. By contrast, for familiar messages there was no significant difference in performance on the time estimation task for the two levels of linguistic redundancy. Possible reasons for the difference in attention requirements for comprehension compared to recognition will be discussed. [Work supported by NASA Grant NGL 05-046-002 to the San Jose State University Foundation.]
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call