A useful test of hazard perception is undoubtedly desirable and should be able to distinguish reliably between drivers who differ in ability (McKenna and Crick, 1991; McKenna and Horswill, 1997). With a number of on-line measures taken during the viewing of hazard perception video clips, the authors have previously reported that differences between drivers can be masked by high variance (Underwood, Crundall and Chapman, 1997). Their purpose here is to describe some of these sources of variance. Drivers vary in their abilities, and hazards vary along a number of dimensions - situational context, abruptness of onset, and projected time-to-contact, as well as hazard object type. In the present study the authors attempt to account for some of the inconsistencies in reports of hazard perception responses by demonstrating differences between drivers and differences between hazard types. (A)
Samenvatting