Vigilance under Highway Driving Conditions

The effect of prolonged driving on the vigilance of Army truck drivers was srudied in cooperation with the American Association of State Highway Officials. A decrement in performance was predicted on the basis of long shifts, a highly repetitive cask, reported physical fatigue, and vehicle noise and vibration. Ss were 42 enlisted Army truck drivers (Men = 23 yr.). Ss drove large, heavily loaded commercial trucks repeatedly around experimental highway driving loops in 9-hr. shifts. One shifc consisted of 7 hr. of actual driving time sequenced in 6 driving periods of 90, 90, 90, 60, 45, and 30 min. Four 15-min. rest pauses and one 40-min. meal break were interspersed among driving periods. Ss monitored a visual discrimination task continuously during driving. Visual display units of 15 lights arranged in a circular fashion were mounted on truck dashboards. Signals appearing in 6 "critical" (red) panels were responded to by means of a foot pedal. Signals appearing in the 9 "noncritical" (white) panels were to be ignored. Critical signals were programmed at the rate of 30 per hour; inter-signal intervals ranging from 5 to 390 sec. occurred in random order. Non-critical signals appeared at the rate of 90 per hour. Signals were oP 1-sec. duration. The absolute level of vigilance was high. Critical signal detections averaged 83% for all driving periods combined. False detections averaged only 4% for all driving periods. There was no between-period performance decrement ( F = 1.21; df = 5/200; P > .25). N o measure of within-period decrement was possible. The absolute detection levels indicated, however, that any decrement that may have occurred within periods was relatively small. Individual differences in critical detection scores were wide (Range = 21 to 100%) and highly reliable (odd versus even driving periods, roc = .93; df = 40; P < .01). The range of individual scores became greater as the vigil progressed, resulting in a significant increase in inter-subject variance from Driving period 1 to 6 ( t = 3.17; df = 41; P < .01). In view of past research, one possible explanation for the present lack of decrement is the relative complexity of the task. Performance decrements have been found repeatedly for simple (single signal source) vigilance tasks (Baker, 1959), but the evidence for decrement in complex (multiple signal source) vigilance tasks is scant (Frankmann & Adams, 1962 ) . Also, the experimental task may have served as a stimulant co rather than a measure of vigilance performance. REFERENCES BAKER, C. H. Towards a theory of vigilance. Canad. I . Psychol., 1959, 13, 35-42. FRANKMAW, J. A,, & ADAMS, J. A. Theories of vigilance. Psychol. Bull., 1962, 59, 257-272. Accepted Decernbw 10, 1962.