The vigilance decrement: its first 75 years.

The first major laboratory studies of vigilance by Mackworth in 1948 and later revealed a decline in the probability of detecting brief targets as the time on task increases. Whether referred to as a vigilance decrement or something else (e.g., a failure of sustained attention), because such failures have great applied significance (e.g., in road safety, radiology, air-traffic control, civil defense, etc.), understanding the vigilance decrement and discovering ways to avoid it are important goals for psychological science. The purpose of this historical review is to provide a picture of the extensive scientific literature exploring the nature(s) of the vigilance decrement, with an emphasis, but not exclusionary focus, on the signal detection theory framework. Beginning in the early 1960s, researchers started to interpret this decline in target detections using signal detection theory, wherein a decrease in detections can be attributed to a decrease in sensitivity of the observer to the difference between targets and non-targets, a conservative shift in the observer's response criterion, or, of course, both. Some early investigators suggested that which of these two causes of the decline in detections is operating may depend on the rate at which events (targets and non-targets combined) are presented: When the event rate is slow, criterion shifts dominate detection failures, whereas declines in sensitivity become more pronounced as event rates increase. Nevertheless, the contribution of sensitivity declines has been recently challenged. One source of the challenge is the relatively low false-alarm rate in so many studies on the vigilance decrement. Another is the possibility that for a variety of reasons, the observer in a relatively long vigil may stop attending to the source of the task-relevant signals. Some recommendations are offered based on our reading of the ~75 years of vigilance research.