| Abstract: | Sonar operators are confronted with a watchstanding task that demands high levels of vigilance for the appearance of weak or transitory signals. Maintaining vigilance is difficult because of very low target signal rates and an open loop system with (usually) no performance feedback. Four experiments were conducted to see whether operator vigilance, as reflected by target detection latency, could be enhanced through signal injection and performance feedback. In each of these experiments, target detection performance was markedly enhanced. The effects were operationally and statistically significant and generally increased with time on watch. The beneficial effects were shown not to be simply a function of increased signal rate due to signal injection. Analysis using a signal detection theory model showed that the target reporting threshold dropped under the experimental treatment and detection efficiency increased. In addition, it was shown that the subjects spent significantly more time observing the search display with signal injection and feedback. On the negative side, there was a modest increase in false alarms which was judged to be tolerable in view of the marked reduction in target detection times. Most false alarms were quickly recognized and reported as such. These beneficial effects were confirmed in a fifth experiment using trained sonar operators as subjects and prototype displays of an advanced sonar system. |
