Methods and Metrics of Voice Communications
Kanki & Prinzo (1996) · Coding cockpit and radio talk
Methods and Metrics of Voice Communications is the proceedings of a two-day workshop in San Antonio run by the FAA, NASA Ames and the US Air Force's Armstrong Laboratory, edited by Barbara Kanki and Veronika Prinzo. Voice here means spoken communication: pilot and controller radio talk, talk between crew, and the acoustics of speech. Its preface says the workshop was meant to further 'the collective understanding of the methods and metrics used in the study of natural language', and Prinzo opens with a problem of definition: within aviation 'communication error' often means a loss of separation attributed to communication, while communication researchers use it more broadly for any action 'based on faulty communication'.
Part 1 holds presentations on discourse and acoustic processes, Part 2 demonstrations of software for coding talk, and Part 3 supplementary material and reprints. Connell analyses 1993 reports to NASA's Aviation Safety Reporting System (ASRS); Bürki-Cohen reports preliminary results from 24 airline pilots on how controllers should say numbers.
Schiflett cites Foushee (1984) for the observation that most team research 'ignored communications as a process variable', and plans to extend the speech matrices of Kanki and Foushee (1989) and Kanki, Lozito and Foushee (1989) beyond two-person crews. Kanki casts communication as a group process mediating between inputs and team performance, its patterns read as symptoms of problems or as strategies, and draws two lessons: 'there are tradeoffs for every method', and the right one depends on the question. Veinott and Irwin make this concrete from a simulation with 22 DC-9 and MD-88 crews: training four coders on 14 speech codes took about nine months, and a minute of tape took about 15 minutes to transcribe, 30 in the abnormal phase. Herbert Clark's keynote, reprinted from 1993, argues that problems in talk are 'really joint problems'.
Among the reprints, Helmreich's 1994 analysis of Avianca Flight 052 finds a number of the crew's status calls to air traffic control 'phrased in a tentative manner', and reads the crash as a system accident, adding national culture with the caveat that 'the hypothesis cannot be proved'. Kanki's study of a launch incident reports an abort call never successfully communicated to the test conductor, and 7% of transmissions crossing organisational boundaries.
The volume does not name psychological safety. Connell relays an ASRS analysts' summary (Chappell, 1994) that includes the 'Reluctance of flight crews to question a controller', especially one who 'sounded rushed, angry, or overloaded', and Helmreich suggests the need to maintain group harmony 'may have inhibited crewmembers from presenting their concerns and suggestions'. The bridge is ours: we read Veinott and Irwin's coding of who asks, who answers and which questions go unanswered as an early way of observing speaking up, though no scheme here was built to measure it.
Save this paper to a bibliography, keep notes on it, and export citations. Create a free account →
Related psychsafety.com articles (5)
- Crew Resource Mgmt
- Reducing Power Gradients
- Human Error & PS
- PS in Aviation
- PS in Aviation (Special Edition)
Related papers (3)
- Helmreich, Merritt & Wilhelm (1999): The Evolution of Crew Resource Management Training in Commercial Aviation
- Foushee (1984): Dyads and Triads at 35,000 Feet: Factors Affecting Group Process and Aircrew Performance
- Kanki & Foushee (1989): Communication as Group Process Mediator of Aircrew Performance
