Voice-Communication Testing Using Naive and Experienced Communicators.
NAVAL RESEARCH LAB WASHINGTON DC
Pagination or Media Count:
Subjects with and without communication experience evaluated a variety of digital voice processors with data rates from 800 to 32,000 bits per second. Two conversational tests were used one with a restricted vocabulary and one with an open vocabulary. Ratings of the voice processors were similar in all conditions, but the experienced communicators used more compensatory behaviors when talking over the poor voice systems. The inexperienced communicators developed similar compensations but used them less often. The results suggest that ratings by the subjects are reliable measures of the performance of voice processors. Objective measures of talking performance are less reliable but show the same trends. Because the experienced subjects were more tolerant of degradation, the ratings did not discriminate among systems as well as those of the inexperienced subjects. However, both groups showed more errors and longer test times with more degraded systems. Thus the best possible system should be chosen even for experienced users, but the choice may better be made with ratings by inexperienced test subjects, and the testing technique need not necessarily closely resemble the real-world application.
- Human Factors Engineering and Man Machine Systems
- Radio Communications
- Voice Communications