Doctor of Philosophy (Ph.D.)
Degree Granting Department
Civil and Environmental Engineering
Fred Mannering, Ph.D.
Abdul Pinjari, Ph.D.
Panagiotis Anastasopoulos, Ph.D.
Michael Maness, Ph.D.
Mingyang Li, Ph.D.
Lu Lu, Ph.D.
Crash frequency analysis, Hazard based duration modeling, Heterogeneity in means, Ordered response model, Random parameters
The injury severity and frequency of traffic crashes can be further reduced by implementing appropriate traffic safety measures. Several advanced methodologies have been developed in the recent decades to more accurately identify the factors affecting road safety. Such methodologies have helped policy makers formulate appropriate safety measures directed toward increasing seatbelt and helmet use, reducing speed limits in school zones, implementing adolescent driver programs, and so on. However, each methodological approach has its own strengths and weakness and could lead to different inferences. This dissertation assesses and explores a number of recently developed methodologies and their application to traffic safety data.First, this study highlights a potential limitation of mixed generalized ordered response models, which allow random heterogeneity in thresholds and are widely used to model ordered outcomes such as accident injury severity. In the mixed generalized ordered response (MGOR) models, the variances of the random thresholds are implicitly assumed to be in a non-decreasing order, and we investigate the use of negative correlations between random parameters as a variance reduction technique to relax the property of non-decreasing variances of thresholds in MGOR models. Second, this study statistically assesses the safety of new motorcycle riders by following the crash history of newly trained motorcyclists for up to 5 years after they complete their required motorcycle training. Weibull hazard-based duration models are then estimated with gamma heterogeneity, and full random parameters, to determine the factors that affect the duration until motorcyclists’ first crash. It is found, after an initial period, crash risk declines over time as increasing riding experience reduces overall risk. This particular finding points to a critical motorcycle-safety opportunity in that policies and/or training can be modified to attempt to eliminate this initial high-risk period to arrive at a desirable lower initial risk and a declining risk over time as experience makes for safer and more cautious riding. Finally, the dissertation concludes by comparing a bivariate model of injury frequencies (the frequency of no-injury crashes and the frequency of injury crashes on the same roadway sections) with the more commonly used univariate modeling approach. It is found that the superiority of one approach over the other is somewhat inconclusive and requires additional empirical exploration. The dissertation concludes with a summary of findings.
Scholar Commons Citation
Balusu, Suryaprasanna Kumar, "An Assessment and Exploration of Recent Methodological Advances in Safety Data Analysis" (2021). USF Tampa Graduate Theses and Dissertations.