Cognitive organization of roadway scenes, part II

An empirical study of roads inside built-up areas


Gundy,drs. C.M




PDF icon pdf (1.37 MB)

This report describes two experimental studies designed to elucidate road users' cognitive organization of urban roadway scenes. (The present study may be viewed in conjunction with the highly similar study of Gundy (1994), conducted on rural roads.) A sample of urban road locations was stratified by seven road classes, three levels of urbanization, and by the presence (or absence) of a intersection nearby. These classes were: Class 0: dual carriageway arteries closed to slow traffic (70 km/hour) Class 1: dual carriageway arteries closed to slow traffic (50 km/hour) Class 2: single carriageway arteries closed to slow traffic (50 km/hour) Class 3: single carriageway arteries open to slow traffic (50 km/hour) Class 4: residential roads open to slow traffic (50 km/hour) Class 5: residential roads open to slow traffic (30 km/hour) Class 6: ‘woonerf' residential roads open to slow traffic (< 15 km/hour). These locations were then photographed from the viewpoint of a driver, and roadside characteristics were registered. The presence of other traffic was avoided as much as possible. A selection of 94 photographs were presented to approximately 25 volunteer subjects per experimental task. (Subjects were chosen such that their age and sex distribution roughly matched that of the Dutch driving popula tion.) In the first experiment, subjects were asked to sort these photographs onto ‘piles' of photographs, placing ‘similar' photographs together, and placing ‘dissimilar' photographs apart. These piles were intended to be ‘meaningful' and ‘useful' to the subjects (as determined by the subjects themselves) in their role as automobile drivers. The sorting data was collected into similarity matrices and analyzed by means of Multi Dimensional Scaling and Analysis of Variance. The findings were quite clear. Namely, subjective similarity judgements were almost entirely ‘explained' by the seven categories mentioned above. In a following study, other subjects were asked to estimate a safe driving speed and the chance of encountering ‘slow' traffic for each of the 94 photographs investigated in the previous study. The results were again analyzed by means of Analysis of Variance, with results clearly mirroring those of the first study. Surprisingly, in contrast to the findings of Gundy's (1994) investigation of rural roads, the presence (or absence) of intersections played only a rather negligible role in the subjects' judgement. Traffic safety implications and possibilities for future research are also considered. More concretely, it is tentatively indicated that there should be essentially three types of urban roads: -high-speed arteries where slow traffic is prohibited, -specially designed residential areas, where all forms of traffic are allowed, yet only (very) low speeds are possible, -and transitional type(s), intermediate to the previous two types

Print this page

Report number



30 + 13


SWOV, Leidschendam