The relation between crash rate, speed and road features on 80 km/h roads; a cross-sectional study. This exploratory study investigates the relation between crash rate, speed, and several features of 80 km/h roads. We carried out this study with driving speed and crash data available to us from another study (SWOV report R-2003-27) and road features that we could relatively easily retrieve. We analysed this data using a cross sectional design in which we examined the relation between various driving speed characteristics and road features, and crash rates. The crash and speed data were derived from 32 roads in the Province of Friesland, for the 1997-2003 period. They were brought together and analysed. We also used data on the road width and junction density, i.e. the number of exits per kilometre. The speed data consisted of the average speed, the speed distribution i.e. the standard deviation of driving speeds, the percentage of offenders (> 87 km/h), and the percentage greatly exceeding the limit (>117 km/h). The crash rate was defined as 'the number of crashes per vehicle kilometre'. We looked at the total crash rate, including material damage only crashes, as well as the injury crash rate i.e. casualty crashes. Practically all analyses were carried out using linear regression analysis. The linear relation between different variables can be described with this technique. The analyses showed that the total crash rate on these limited access 80 km/h roads is higher on roads with a greater junction density, on narrower roads, and on roads with a lower average speed. Combined with these factors, the crash rate is also higher on roads with a higher percentage of speeding offenders. The injury crash rate was especially higher on narrow roads and on roads with a lower average speed. Combined with these factors, the injury crash rate is also higher on roads with greater speed distribution. Contrary to results of other studies, this study showed no direct relation between road features, such as road width, and speed. The mutual relation of speed variables, however, is strong: higher average speed goes together with greater speed distribution and higher percentages of offenders greatly exceeding the limit. Road features as well as speed go together in their own way with variations in crash rates. They are thus both important for gaining insight in how large crash rates on roads are. Based on these results, we conclude that there may be more aspects of the road design and environment, and possibly also different road user types that play a role, than those that we analysed in this exploratory study. We also make recommendations for future studies.
Samenvatting