Journal
TRANSPORTATION RESEARCH PART B-METHODOLOGICAL
Volume 49, Issue -, Pages 39-54Publisher
PERGAMON-ELSEVIER SCIENCE LTD
DOI: 10.1016/j.trb.2013.01.002
Keywords
Adolescent crashes; Graduated driver licensing; Heterogeneity model; Gaussian mixture; Bayesian inference; Permutation sampler
Ask authors/readers for more resources
One of the key aspects of graduated driver licensing programs is the new-driver experience gained in the presence of a guardian (a person providing mandatory supervision from the passenger seat). However, the effect that this guardian-supervising practice has on adolescent drivers' crash-injury severity (should a crash occur) is not well understood. This paper seeks to provide insights into the injury-prevention effectiveness of guardian supervision by developing an appropriate econometric structure to account for the complex interactions that are likely to occur in the study of the heterogeneous effects of guardian supervision on crash-injury severities. As opposed to conventional heterogeneity models with standard distributional assumptions, this paper deals with the heterogeneous effects by accounting for the possible multivariate characteristics of parameter distributions in addition to allowing for multimodality, skewness and kurtosis. A Markov Chain Monte Carlo (MCMC) algorithm is developed for estimation and the permutation sampler proposed by Fruhwirth-Schnatter (2001) is extended for model identification. The econometric analysis shows the presence of two distinct driving environments (defined by roadway geometric and traffic conditions). Model estimation results show that, in both of these driving environments, the presence of guardian supervision reduces the crash-injury severity, but in interestingly different ways. Based on the findings of this research, a case could easily be made for extending the time-requirement for guardian supervision in current graduated driver license programs. (C) 2013 Elsevier Ltd. All rights reserved.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available