Numerous vigilance researchers have questioned the wisdom of generalizing laboratory results to real-world tasks (Adams, 1987; Alluisi, Coates, and Morgan, 1977; Elliott, 1960; Kibler, 1965; Mackie, 1987; Thackray, Bailey, and Touchstone, 1977). Their concerns arise mainly from three observations:
1. laboratory vigilance tasks rarely resemble their real-world counterparts;
2. when real-world tasks are adequately simulated, the results often differ from laboratory findings; and
3. there are "few, if any, troublesome vigilance decrements in the operational tasks of the real world" (Adams, 1987 p. 737).
Although admitting that a large body of important psychological knowledge has accrued from laboratory vigilance research, Wiener (1987) cautioned that vigilance must be studied in more complex and operationally valid environments. He noted that real-world vigilance tasks usually involve highly trained and experienced individuals working long hours on critical tasks with relatively low signal rates.
Unfortunately, conducting research in such environments, without impeding or disrupting ongoing activity, is often difficult. Measurement techniques need to be nonintrusive, experimental protocols must be consistent with the demands of the workplace, and the results have to be relevant (and helpful) to the individuals being studied.
We have had the opportunity to study such a real-world vigilance task (not a simulation) under daily operational conditions, with control over several factors. One function of North American Aerospace Defence (NORAD) in North Bay, Ontario, is to identify all aircraft entering Canadian airspace. This is accomplished by visually detecting potential intruders from information presented on display consoles and correlating the detected targets with logged aircraft flight plans. If the targets do not correlate with logged flight data, then interceptor aircraft may be scrambled to visually identify the transgressor.
Because the entire process depends on rapid initial detection by the surveillance operators, it is important to have a good understanding of the factors influencing detection performance in this operational environment.
The Air Surveillance Task and Environment
Canadian NORAD Region's air surveillance has been continuously maintained for 35 years and involves monitoring returns from an array of radars covering Canada's eastern, western, and northern territories. The radar information is data-linked to Canadian NORAD Region headquarters in North Bay.
Canada's area of radar coverage is divided into two geographic regions, Canada East and Canada West, the surveillance of which is given to two squadrons that physically, functionally, and operationally are independent. The air picture is updated every 12 s by NORAD computers, which must fuse and correlate all radar data to eliminate redundant echoes reflected from adjacent radar sites. Each geographic region is then subdivided into zones that can be displayed on individual surveillance consoles. The display consoles are round, monochrome (green phosphor), and vector scanned. Air surveillance operators do not monitor returns from individual radars; they monitor fused information of geographic zones serviced by multiple radars. To represent aircraft movement, the seven most recent (fused) radar returns are sequentially displayed every 2 s. This leaves a blip trail, or track, showing an aircraft's heading and speed.
The surveillance operator's task is to detect these tracks and direct the computer to tag them and assign them track numbers. The operators are not responsible for identifying the tracks, only for detecting them. Depending on the type of aircraft and its intention, a track can have one of two "signatures." Aircraft equipped with transponders emit status and identification signals that are received by the radars, interpreted by NORAD computers, and displayed on the consoles with a diamond symbol. …