A note on modeling pedestrian-injury severity in motor-vehicle crashes with the mixed logit model.
ABSTRACT Pedestrian-injury severity has been traditionally modeled with approaches that have assumed that the effect of each variable is fixed across injury observations. This assumption ignores possible unobserved heterogeneity which is likely to be particularly important in pedestrian injuries because unobserved physical health, strength, and behavior may significantly affect the pedestrians' ability to absorb collision forces. To address such unobserved heterogeneity, this research applies a mixed logit model to analyze pedestrian-injury severity in pedestrian-vehicle crashes. Using police-reported collision data from 1997 through 2000 from North Carolina, several factors were found to more than double the average probability of fatal injury for pedestrians in motor-vehicle crashes including: darkness without streetlights (400% increase in fatality probability), vehicle is a truck (370% increase), freeway (330% increase), speeding involved (360% increase), and collisions involving a motorist who had been drinking (250% increase). It was also found that the effect of pedestrian age was normally distributed across observations, and that as pedestrians became older the probability of fatal injury increased substantially. Heterogeneity in the mean of the random parameters for the freeway and pedestrian-solely-at-fault collision indicators was related to pedestrian gender, and heterogeneity in the mean of the random parameters for the traffic-sign and motorist-back-up indicators was related to pedestrian age.
- Journal of the American Dietetic Association 07/2011; 111(7):986-90; author reply 990-3. · 3.80 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: This study identifies and compares the factors that contribute to injury severity on urban freeways and arterials and recommends potential countermeasures to enhance the safety of both facilities. The study makes use of an extensive data set from the State of Florida in the United States. To obtain a more complete picture, this study explores both traditional and nontraditional severity predictors. Some traditional predictors include traffic volume, speed limit, and road surface condition. The nontraditional predictors are comprised of those rarely explored in previous severity studies, including crash distance to the nearest ramp location, detailed vehicle types, and lighting and weather conditions. The analysis was conducted using the ordered and binary probit models, which are well suited for the inherently ordered property of injury severity. An important finding is the significance of the distance of crash to the nearest ramp junction/access point, for which the increase in the distance yielded a severity increase at both facilities. Other significant factors included traffic volume, speed limit, at-fault driver's age, road surface condition, alcohol and drug involvement, and left and right shoulder widths. In comparing both facilities, sport utility vehicles (SUVs) and pickup trucks showed a fatality/severity increase on freeways and a decrease on arterials. Furthermore, the detailed list of variables such as crash time provided pertinent severity trend information that showed that, compared to the other periods, the afternoon peak period had the highest reduction in fatality/severity. Both probit models succeeded in identifying significant severity predictors for each facility. The binary probit model outperformed the ordered probit model based on the higher elasticities (marginal effects) for the fatality/severity probability change, as well as the goodness of fit. As such, this study provides the guidelines for assessing the impact of important roadway and traffic characteristics on crash injury severity along freeways and arterials.Traffic injury prevention 06/2011; 12(3):223-34.
- [Show abstract] [Hide abstract]
ABSTRACT: The current study focuses on the propensity of drivers to engage in crash avoidance maneuvers in relation to driver attributes, critical events, crash characteristics, vehicles involved, road characteristics, and environmental conditions. The importance of avoidance maneuvers derives from the key role of proactive and state-aware road users within the concept of sustainable safety systems, as well as from the key role of effective corrective maneuvers in the success of automated in-vehicle warning and driver assistance systems. The analysis is conducted by means of a mixed logit model that represents the selection among 5 emergency lateral and speed control maneuvers (i.e., "no avoidance maneuvers," "braking," "steering," "braking and steering," and "other maneuvers) while accommodating correlations across maneuvers and heteroscedasticity. Data for the analysis were retrieved from the General Estimates System (GES) crash database for the year 2009 by considering drivers for which crash avoidance maneuvers are known. The results show that (1) the nature of the critical event that made the crash imminent greatly influences the choice of crash avoidance maneuvers, (2) women and elderly have a relatively lower propensity to conduct crash avoidance maneuvers, (3) drowsiness and fatigue have a greater negative marginal effect on the tendency to engage in crash avoidance maneuvers than alcohol and drug consumption, (4) difficult road conditions increase the propensity to perform crash avoidance maneuvers, and (5) visual obstruction and artificial illumination decrease the probability to carry out crash avoidance maneuvers. The results emphasize the need for public awareness campaigns to promote safe driving style for senior drivers and warning about the risks of driving under fatigue and distraction being comparable to the risks of driving under the influence of alcohol and drugs. Moreover, the results suggest the need to educate drivers about hazard perception, designing a forgiving infrastructure within a sustainable safety systems, and rethinking in-vehicle collision warning systems. Future research should address the effectiveness of crash avoidance maneuvers and joint modeling of maneuver selection and crash severity.Traffic injury prevention 05/2012; 13(3):315-26.