The effects of transmission delay in four-wire teleconferencing

The effects of transmission delay upon the performance of a three-party teleconference were investigated using a problem-oriented task. The teleconference was simulated in the laboratory using 4-wire telephone sets interconnected to form a three-party conference network without echo sources or echo control devices. The two experimental conditions were characterized by (a) a network whose three legs provided transmission delays of 600-milliseconds, 300-milliseconds and no delay, and (b) a network each of whose three legs provided no transmission delay. It was found that: (i) time to complete the experimental task was 28 percent greater in the delay condition than in the no-delay condition, (ii) the error rate was less in the delay condition than in the no-delay condition, (iii) time-per-trial decreased with successive trials in both delay conditions; time-per-trial was less in the no-delay condition than in the delay condition, and (iv) no chairmanship pattern developed as a result of time delay in the network. Moreover, not one of the subjects reported having observed the existence of delay in the voice path. There were, however, more complaints of “talking together” in the delay condition.