|
Abstract
|
The authors have retracted this report after they discovered that a fatal error was made in the logistic regression model. More information is available at:
DOI: 10.1016/j.jth.2014.09.005
The purpose of this study was to analyze the relationship between the likelihood of bicycle collisions and time of day, traffic volume, road surface conditions, legal speed limit, and weather.
Data from the Saskatchewan Traffic Accident Information System (TAIS) was used. Descriptive statistics, mapping and logistic regression analysis were conducted to examine factors associated with bicycle collisions.
Legal speed limits greater than 70 km/h were shown to be more dangerous when compared to legal speed limits less than 50 km/h, which showed an 84% reduced probability of collisions (OR=0.26, CI: 0.11 to 0.54). Compared to morning cycling, night time cycling was associated with 2.13 times the risk of collision (95% CI: 1.28 to 3.82). Hazardous road surface conditions, such as snow and ice, increased the likelihood of a collision by 12.13 times (95% CI: 8.13 to 19.98). Contrary to previous research, our analysis showed that collisions do not occur more frequently at intersections.
The results from this study demonstrated that road conditions and time of day play an important role in cycling collisions. Transportation and public health policies such as regular maintenance of roadways for cyclists and more bicycle scale street lights could address these factors at relatively low cost. |