ROYAL NAVAL PERSONNEL RESEARCH COMMITTEE LONDON (ENGLAND)
Two separate groups of 16 naval ratings were asked to detect pure tone signals presented in noise. Signal probability was 0.1. One group of Ss performed with the assistance of a simulated computer functioning at 75 accuracy the other group performed without computer assistance. Assistance was provided by 6 visual indications of confidence regarding the signal. Ss responded with 4 levels of confidence regarding the signal. Both computer and man used confidence ratings ranging from certain nono signal to certain yessignal. Ss were given practice at the task and then tested for 75 min per day for 8 days. Computer assistance was found to increase the proportion of correct detections made. There was some evidence that detections of the unassisted group declined over the test and that assistance reduced this decline.