The overall objective of this project was to evaluate roadway and operational factors influencing the high injury and fatality rates on six-lane divided roadways in Florida. To accomplish this objective, data from the FDOT crash database, the Roadway Characteristics Inventory, FDOT videologs, and field studies were analysed. Although four-lane roadways had more crashes than six-lane roadways in terms of percentages, six-lane roadways had higher crash, fatality, and injury rates based on many environmental and geometric factors. The Spearman and chi-square tests showed that, based on a number of geometric and traffic variables, the distribution of crashes on four-lane and six-lane highways were not significantly different. Zero-inflated negative binomial models were used to model crash rates according to three levels of severity: (1) fatal and severe injury, (2) non-incapacitating and possible injury, and (3) property damage only. Increases in the number of signals per miles and the number of driveways per mile increased the crash rate. Increases in inside shoulder width, horizontal degree of curvature, outside shoulder width, median width, and surface width all reduced the crash rate. Based on the results of this research, the researchers strongly endorse FDOT's efforts to enforce its standards, especially with regard to median width, shoulder width, and access management. (Author/publisher)
Abstract