Training for Vigilance: A Comparison of Different Techniques

Abstract

In a 40-minute vigilance session, 72 subjects inspected a series of displays, each of which consisted of a row of 6 small disks, for the occasional presence of a disk 17 percent greater in area than the remainder. The possibility of improving the generally low levels of performance observed with this task by special training was studied by pre-exposing subjects to a session of similar length with either knowledge of results (KR), one of 3 kinds of cueing of signal occurrence, a mixed KR/cueing program, or no task information. No differential effect of the various training techniques was found, but both the efficiency with which signals were discriminated, and the degree of caution exercised in reporting their occurrence increased during the experiment. It is concluded that greater understanding of the factors affecting signal detectability and decision-criteria in vigilance tasks is required before an appropriate method of training can be devised.