Survival analysis: Pedestrian risk exposure at signalized intersections

Pedestrian behaviour was observed at seven selected intersections in Delhi, India. Data collection occurred at these intersections by placing a video camera at each zebra crossing. Pedestrian crossing behaviour was then obtained from careful review of the videotapes. Pedestrian crossing behaviour was analyzed using survival analysis statistical methodology. The analyses produce Kaplan–Meier survival curves for waiting time prior to crossing unsafely, separately for males and females, and for each intersection. Mean observed waiting time and model-based waiting time of 90% of pedestrians were then studied. Mean waiting time of females are 27% more than for males, while the waiting time of 90% of female pedestrians are 44% more than the corresponding number for males. The probability for a pedestrian to cross the road, when it is unsafe, i.e. motor vehicles still have green or yellow, varies with waiting time. People do not want to wait too long to cross streets. As signal waiting time increases, pedestrians get impatient and violate the traffic signal. This violation places them at increased risk of being struck by a motor vehicle. Thus, reducing the waiting time for pedestrians are likely to decrease the probability of pedestrian crossers being hit by a motor vehicle.