Objective: Vehicle crashes that involve pedestrians at intersections have been reported occasionally. Pedestrian injury severity in these crashes is significantly related to driver and pedestrian attributes, vehicle characteristics, and the geometry of intersections. Identifying factors associated with pedestrian injury severity (PIS) is critical for reducing crashes and improving safety. For developing the proposed probit models, drivers involved in crashes are classified into 3 groups: young drivers (16 ≤ age ≤ 24), middle-aged drivers (25 ≤ age ≤ 64), and older drivers (age ≥ 65). This study determines that PIS is significantly but differently affected by these grouped drivers with different sets of explanatory variables. Methods: A total of 2,614 crash records (2011–2012) at intersections in Cook County, Illinois, were collected. An ordered probit modeling approach was employed to develop the proposed model and examine factors influencing PIS. The likelihood ratio test was used to assess model performance. Elasticity analysis was conducted to interpret the marginal effect of contributing factors on PIS associated with different driver groups by age. Results: The results show that 4 independent variables, including pedestrian age, vehicle type, point of first contact, and weather condition, significantly affect PIS at intersections for all drivers. Two additional independent variables (i.e., number of vehicles and traffic type) affect PIS for young and middle-aged drivers, and 2 other variables (i.e., divided type and hit-and-run related) are significant to PIS for both young and older drivers. Conclusions: The independent variables significant to PIS at intersections for young, middle-aged, and older driver groups were identified and the marginal effect of each variable to the likelihood of PIS were assessed.
In order to identify the possible contributing factors on two-lane rural highways to accident severity, the multinomial logit model was applied to traffic accident data collected from police-reports. A total of 253 accidents, occurred on two-lane rural highways in 2007, were considered in this paper. The dependent variable is divided into three categories, namely fatal accident, injury accident, and property damage only accident, and 13 independent variables were selected in this paper. Of 13 independent variables obtained from police-reported accidents, five were significantly associated with accident severity, namely location, weather, clients' sex, vehicle type, and accident type. The findings show that the multinomial logit model used in this paper provides a better understanding of risk factors related to two-lane rural highways.
In order to deeply understand the relationship between traffic accident frequency and potential influence factors on a freeway, 554 cases of traffic accidents, which occurred on a certain freeway from 2006 to 2008, were analyzed. The homogeneous longitudinal grade was selected as division criterion and the developed section was divided into 150 segments. In order to analyze the relationship between traffic accident frequency and various factors, twelve independent variables were selected and a traffic accident frequency prediction model was established, based on traditional negative binomial (NB) model and random effects negative binomial (RENB) model respectively.. The results show that RENB model is better than NB model by comparing the results of goodness-of-fit. For RENB model, four independent variables, including slope grade, road width, subtraction of slope grade between adjacent segments, and annual traffic volume, have significant impact on accident frequency on a freeway. The marginal effects of four independent variables in RENB model are also analyzed.