A study of the relationship between cognitive road classification and actual driving behaviour is described. The experiment was divided into a picture sorting task (based on road classification in the Netherlands) and a driving simulator task using conventional and self-explaining roads. The driving simulation was conducted without other vehicles, traffic signs or sharp bends. All roads within a given category were adjusted to have the same width, with road width decreased as road quality decreased. Forty-eight participants 23-45 years of age including males and females took part in the study. All had possessed driving licenses for more than 5 years and drove more than 10,000km/year. In the simulation study participants drove faster as road quality increased and in road categories B and D, participants drove significantly faster on self-explaining than on conventional roads. All participants were exposed three times to each set of roads and their driving speeds increased with repetition of the study, although the differential speed between road categories remained the same. Participants reported that the presence of houses or trees next to roads moderated their speed, whereas speed was increased on straight roads and dual carriageways. Road categorisation was improved by self-explaining road design.
Abstract