This paper reports research aiming to test the relationship between Driver Behaviour Inventory (DBI) scales and measures of driver simulation performance related to attention and risk-taking. Four samples were run, some of which performed more than one task, so that the data were analysed in six studies. Each driver viewed a three-dimensional display of the road ahead, and controlled the 'car' with a steering wheel, brake, and accelerator pedals. The state of the controls and 'car' were recorded every 500 ms. The simulator was programmed to investigate three types of driving situation: (1) open-road tasks; (2) following tasks; and (3) overtaking tasks. A run typically lasted for 10 min, and was over a 'road' that was an unpredictable mixture of straight sections and bends of varying curvature. The data show that all three DBI scales predict some aspects of performance, but only under specific circumstances. Aggression was considerably correlated with risk-taking only in the overtaking studies. Aggressive drivers seemed to face special risks in overtaking situations. Dislike of driving had less clear safety implications. More alert drivers paid better attention to hazards. The results imply that a variety of mechanisms influence driver behaviour, with their dangerous effects apparently greater than their beneficial effects. However, the simulator results may not generalise to real driving, and their relatively mild experimental stresses may have been unrealistic.
Abstract