Effect of prolonged driving on overtaking criteria.

During the first and last 3 hours of a 12-hour spell of driving by 6 subjects the experimenter recorded the number of overtaking manoeuvres which were begun with what he judged to be an unduly high level of risk. When the data were corrected for inter-session differences in frequency of overtaking, risky manoeuvres were found to be initiated 50 per cent more often during the last 3 hours of the spell than during the first. This result conflicts with Potts' (1951) finding that lsquo;near accidents’ decrease with time on task, but it supports earlier conclusions on the superiority of assessing fatigue effects subjectively rather than objectively.