Effects of parallelogram-shaped pavement markings on vehicle speed and safety of pedestrian crosswalks on urban roads in China

The primary objective of this study was to evaluate the effects of parallelogram-shaped pavement markings on vehicle speed and crashes in the vicinity of urban pedestrian crosswalks. The research team measured speed data at twelve sites, and crash data at eleven sites. Observational cross-sectional studies were conducted to identify if the effects of parallelogram-shaped pavement markings on vehicle speeds and speed violations were statistically significant. The results showed that parallelogram-shaped pavement markings significantly reduced vehicle speeds and speed violations in the vicinity of pedestrian crosswalks. More specifically, the speed reduction effects varied from 1.89 km/h to 4.41 km/h with an average of 3.79 km/h. The reduction in the 85th percentile speed varied from 0.81 km/h to 5.34 km/h with an average of 4.19 km/h. Odds ratios (OR) showed that the parallelogram-shaped pavement markings had effects of a 7.1% reduction in the mean speed and a 6.9% reduction in the 85th percentile speed at the pedestrian crosswalks. The reduction of proportion of drivers exceeding the speed limit varied from 8.64% to 14.15% with an average of 11.03%. The results of the crash data analysis suggested that the use of parallelogram-shaped pavement markings reduced both the frequency and severity of crashes at pedestrian crosswalks. The parallelogram-shaped pavement markings had a significant effect on reducing the vehicle-pedestrian crashes. Two crash prediction models were developed for vehicle-pedestrian crashes and rear-end crashes. According to the crash models, the presence of parallelogram-shaped pavement markings reduced vehicle-pedestrian crashes at pedestrian crosswalks by 24.87% with a 95% confidence interval of [10.06-30.78%]. However, the model results also showed that the presence of parallelogram-shaped pavement markings increased rear-end crashes at pedestrian crosswalks by 5.4% with a 95% confidence interval of [0-11.2%].